system

Abstract

Provide a system. [Solution] A means of collecting content whose authenticity has been verified by a certification body; A means to verify digital signatures or authentication certificates on collected content; a means for storing the verified authentic content in a database; A means to train a generative AI model based on the content in the database, and means for receiving a request from a user and analyzing the request; means for searching for relevant content in a database based on the parsed request; A system that includes a means for a generative AI model to generate answers based on search results.

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] Advances in modern generative AI technology have made it increasingly difficult to verify the authenticity of online content. This has led to the frequent spread of fake news and hoaxes, negatively impacting society. If this problem is left unaddressed, it will become extremely difficult to identify reliable information, undermining the accuracy and reliability of the information. Therefore, a system is needed in which generative AI can train only on content certified by trusted certification organizations and provide appropriate answers. [Means for solving the problem]

[0005] The present invention provides a system including: means for collecting content whose authenticity has been confirmed by a certification organization; means for verifying the digital signature or certification certificate of the collected content; means for storing the confirmed authentic content in a database; means for training a generative AI model based on the content in the database; means for receiving requests from users and analyzing the requests; means for searching for related content in the database based on the analyzed requests; and means for the generative AI model to generate answers based on the search results. This system can provide users with reliable information and prevent the spread of fake news and hoaxes. Furthermore, by including means for analyzing and classifying the collected content, removing unnecessary and duplicate data, and means for quality checking the generated answers, more accurate information can be provided. Furthermore, by using a public API and secure data transfer means for periodically obtaining certified content from the certification organization, the database can always be populated with the latest, reliable information.

[0006] A "certification body" is an organization established by a public or private organization to provide reliable information and to verify and certify the authenticity and quality of content.

[0007] "Content" is any form of digital information that provides information or knowledge, such as text, images, audio, video, or data.

[0008] A "digital signature" is an electronically generated identification that is used to ensure the integrity of content and the authenticity of its origin.

[0009] "Certification" means a document or symbol, either electronic or physical, that is evidence that a certification body has formally recognized that particular content is trustworthy.

[0010] A "database" is a collection of systematically organized digital information, a storage system structured to allow efficient searching and retrieval.

[0011] A "generative AI model" is a type of artificial intelligence that uses machine learning and natural language processing technologies to learn from large amounts of data and generate new information and sentences.

[0012] "User" means an individual or entity that uses a computer system or service to collect, manipulate, or view information.

[0013] A "request" is a request for information or a question that a user sends to the system.

[0014] "Search" is the operation of locating information in a database based on specific conditions.

[0015] An "answer" is information provided by the system in response to a user request, which may be text or other form of information generated by a generative AI model.

[0016] "Metadata" is data that indicates information about the data itself, and includes the creation date and time of the data, the creator, the content category, and the like.

[0017] A "public API" is an application programming interface that is open to the public for access by external systems and provides a means to access specific functions or data.

[0018] "Secure data transfer" means a communications technology designed to protect data transmission and reception from unauthorized access or tampering by third parties. [Brief explanation of the drawings]

[0019] [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. 3 is a sequence diagram illustrating a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0020] 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.

[0021] First, the terms used in the following description will be explained.

[0022] 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, a 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), and an APU (Accelerated Processing Unit).

[0023] 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.

[0024] 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.

[0025] 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.

[0026] 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."

[0027] [First embodiment]

[0028] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0029] 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.

[0030] 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).

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] 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.

[0037] 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.

[0038] 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.

[0039] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. 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."

[0040] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body. The purpose of this system is to provide reliable information and prevent the spread of fake news and hoaxes. A specific embodiment of this system is described below.

[0041] System Overview

[0042] This system involves a server, a terminal, and a user working together. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[0043] Program processing

[0044] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[0045] server

[0046] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[0047] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[0048] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[0049] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. Based on this, the server searches for relevant content in the database and extracts reliable information.

[0050] Based on the extracted information, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. Once the final answer is determined, the server sends it to the device.

[0051] Terminal

[0052] The terminal acts as the user's interface: the user enters a question or request, which the terminal sends to the server, and upon receiving the answer from the server, the terminal displays the answer to the user.

[0053] User

[0054] Users access the system through terminals, which they type into the terminal questions or requests for information, which then transmit the requests to the server, and the server's responses are displayed on the terminal.

[0055] Specific examples

[0056] Example 1: Medical information question

[0057] User Input

[0058] The user inputs the question "How can I relieve cold symptoms?" into the terminal.

[0059] Server Processing

[0060] The server analyzes the question and searches a database for relevant and reliable medical information. Based on information from certified medical institutions, a generative AI model generates an appropriate answer.

[0061] answer

[0062] The server sends an answer such as, "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays the answer to the user.

[0063] This invention allows users to obtain reliable information and prevents the spread of fake news and rumors.

[0064] The processing flow will be explained below.

[0065] Step 1: Please give me some.

[0066] Example 1

[0067] Next, a description will be given of 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."

[0068] Today, the internet is flooded with unreliable information and fake news, making it difficult for users to quickly obtain reliable information. This problem is particularly severe in specialized fields such as medicine and law, where false information can have a significant impact. Furthermore, conventional systems do not fully establish procedures for verifying reliability, formatting data, and checking the quality of generated answers, meaning the reliability of the information provided to users cannot be guaranteed. To address these issues, a generative AI system is needed that can provide reliable information and prevent the spread of fake news and rumours.

[0069] 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.

[0070] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the verified content in a database, means for analyzing the stored content, classifying it according to format and content, deleting unnecessary data and duplicate data, and formatting it into a dataset, means for training a generative AI model based on the formatted dataset in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed request, means for the generative AI model to generate an answer based on the search results, and means for quality checking the generated answer to check for problems with grammar and content. This allows users to obtain reliable information quickly and accurately.

[0071] A "certification body" is an organization that verifies and certifies the accuracy and reliability of specific content or information with the aim of providing reliable information.

[0072] A "digital signature" is an electronic means of authentication that uses cryptographic technology to verify the identity of the sender of electronic information and whether the information has been tampered with.

[0073] A "certificate of accreditation" is a paper or digital certificate issued by an accreditation body to officially certify the authenticity or accuracy of particular content.

[0074] A "database" is a system for efficiently and systematically storing and managing information and data.

[0075] A "generative AI model" is an artificial intelligence algorithm that learns patterns and relationships from large amounts of data and generates new information and answers.

[0076] A "user request" is a question or request that a user sends to the system for specific information.

[0077] "Natural language processing" is the technology that enables computers to understand, analyze, and generate human language.

[0078] "Quality check" is an inspection process to check whether there are any problems with the grammar or content of the generated answer and to provide it to the user in the most optimal form.

[0079] A "public API" is a public application programming interface that provides an interface for external developers to access specified functionality and data.

[0080] A "secure data transfer method" is a data transfer method that uses encryption and authentication to prevent information leaks and unauthorized access when sending and receiving data.

[0081] This invention is a generative AI system designed to provide reliable information and prevent the spread of fake news and hoaxes. This system operates in cooperation with a server, terminals, and users.

[0082] The server performs processing using the following hardware and software:

[0083] Hardware: High-performance processor, memory, storage, network interface

[0084] Software: public APIs, secure data transfer methods (e.g., TLS), digital signature verification tools (e.g., OpenSSL), databases (e.g., MySQL®), analysis scripts (e.g., Python and the pandas library), generative AI models (e.g., TENSORFLOW®), and natural language processing tools (e.g., Google® Cloud Natural Language API and Grammarly API).

[0085] 1. Content Collection

[0086] The server runs a scheduled task to collect trusted content from trusted organizations and sends HTTP GET requests to designated API endpoints. For example, to retrieve medical information, the server retrieves data from the API of a trusted medical institution.

[0087] 2. Reliability check

[0088] The server verifies the digital signature and authentication certificate of the retrieved content. This process uses "OpenSSL" to verify whether the content signature is correct, for example, by using the "openssl verify" command.

[0089] 3. Saving to the database and formatting

[0090] Once the content is verified as reliable, it is stored in a MySQL database. Python scripts are then run to parse the data, classify it by format and content, and clean it up by removing unnecessary and duplicate data, allowing for efficient searching and learning.

[0091] 4. Training the generative AI model

[0092] The resulting dataset is used to train a generative AI model using the TensorFlow library, and each time new data is added to the system, the model is retrained to improve its accuracy.

[0093] 5. User Request Analysis

[0094] The server receives questions sent by users from their devices and analyzes them using natural language processing technology. The server uses the Google Cloud Natural Language API to extract intent and keywords.

[0095] 6. Extraction of reliable information and generation of answers

[0096] The server searches for relevant content in the database and generates an appropriate answer using a generative AI model. This generative AI model generates an answer based on a "prompt sentence." For example, if a user enters "How can I relieve cold symptoms?", the prompt sentence is set to "Provide appropriate medical information based on the user's question."

[0097] 7. Quality check of answers

[0098] The generated answers are quality checked using the Grammarly API to ensure there are no problems with grammar or content.

[0099] 8. Submitting and Viewing Your Answers

[0100] The final answer is sent from the server to the device as an HTTP response, and the device displays it to the user, who can check the answer through the device's interface.

[0101] Specific examples

[0102] Example: Questions about medical information

[0103] User input:

[0104] The user types into the terminal, "How can I relieve my cold symptoms?"

[0105] Server Action:

[0106] The server analyzes the question using the Google Cloud Natural Language API and extracts keywords (e.g., "cold," "symptoms," "relieve").

[0107] The server searches the database for relevant and reliable medical information.

[0108] The generative AI model uses the prompt "How to relieve cold symptoms" to generate an answer.

[0109] The generated answers are quality checked using the Grammarly API.

[0110] answer:

[0111] The server sends the answer "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays it to the user.

[0112] This invention allows users to obtain reliable information quickly and accurately, and eliminates the influence of fake news and hoaxes.

[0113] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0114] Step 1: Content Collection

[0115] The server collects trusted content from a certification authority. As input, it uses the public API endpoint URL. Specifically, the server runs a scheduled task (e.g., a cron job) to send an HTTP GET request. The output is the retrieved content data.

[0116] Step 2: Authenticity check

[0117] The server verifies the digital signature and authentication certificate of the retrieved content. It uses the retrieved content data and its digital signature as input. Specifically, the server uses the "OpenSSL" library to verify the digital signature. The output is the content, which has been confirmed to be authentic.

[0118] Step 3: Saving to the database and formatting

[0119] The server stores the verified content in a database and formats it. The verified content data is used as input. Specifically, the server stores the data in a MySQL database, runs a Python script to analyze the data, and classifies it according to format and content. It then removes unnecessary and duplicate data and formats it into a dataset. The output is a formatted dataset.

[0120] Step 4: Training the generative AI model

[0121] The server trains the generative AI model using the formatted dataset. The formatted dataset is used as input. Specifically, the server retrains the generative AI model using the TensorFlow library. The output is a trained generative AI model.

[0122] Step 5: Receiving a User Request

[0123] The terminal receives requests from the user. As input, it uses questions or requests that the user types into the terminal. Specifically, the terminal receives input through a user interface and sends it to the server. The output is the user request sent to the server.

[0124] Step 6: Parsing the user request

[0125] The server receives the user request and analyzes it using natural language processing technology. The user request is used as input. Specifically, the server uses the Google Cloud Natural Language API to extract the intent and keywords of the request. The output is the analyzed intent and keywords.

[0126] Step 7: Extract reliable information

[0127] The server searches for relevant content in a database based on the parsed intent and keywords. It uses the parsed intent, keywords, and database as input. Specifically, the server executes SQL queries to extract relevant information. The output is the relevant content.

[0128] Step 8: Answer Generation

[0129] The server uses a generative AI model to generate an appropriate answer based on related content. The server uses related content and the generative AI model as input. Specifically, it inputs a "prompt sentence" into the generative AI model to generate an answer. The output is the generated answer. For example, the prompt sentence could be "Provide appropriate medical information based on the user's question."

[0130] Step 9: Check the quality of your answers

[0131] The server checks the quality of the generated answer. It uses the generated answer as input. Specifically, the server checks the grammar and content using the Grammarly API. The output is a quality-checked answer.

[0132] Step 10: Submit and view your responses

[0133] The server sends the quality-checked answer to the terminal, which displays it to the user. The quality-checked answer is used as input. Specifically, the server generates an HTTP response and sends it to the terminal. The terminal analyzes the response data and displays it in its user interface. The output is the answer displayed to the user.

[0134] (Application example 1)

[0135] Next, a description will be given of Application 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."

[0136] In today's information society, the spread of fake news and inaccurate information has become a serious problem. This increases the risk that users will make incorrect decisions based on unreliable information. In particular, unreliable information can easily spread on web pages and social media, so there is a need for a mechanism to quickly detect this and notify users.

[0137] 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.

[0138] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and authentication certificates of the collected content, means for storing the confirmed reliability content in a database, means for training a generative AI model based on the content in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed requests, means for the generative AI model to generate an answer based on the search results, means for evaluating the reliability of content on web pages and social media that users are viewing in real time, and means for displaying warnings for unreliable content. This allows users to make decisions based on reliable information and prevents the spread of fake news and hoaxes.

[0139] An "accreditation body" is an organization or group whose role is to officially verify and certify the reliability of information.

[0140] "Authenticated content" means information or data that has been formally verified and certified for accuracy and authenticity by an accredited organization.

[0141] A "digital signature" is an electronic certificate that is given to ensure the authenticity and integrity of content.

[0142] A "certificate of authenticity" is an official certification that particular content comes from a trusted source.

[0143] A "database" is a collection of information that stores collected content in an organized manner and can be efficiently searched and used.

[0144] A "generative AI model" is an artificial intelligence model that has the ability to learn patterns and relationships from large amounts of data and generate new information and answers.

[0145] A "request" is an inquiry or request that a user sends to the system for information.

[0146] "Analysis" is the process of extracting the intent and keywords of the user's request and understanding their meaning.

[0147] "Related content" is information or data that is relevant to the user's request.

[0148] An "answer" is appropriate information or a solution generated based on a user request.

[0149] A "web page" is a collection of information or data that can be viewed on the Internet.

[0150] "Social media" refers to online platforms that allow users to share information and communicate over the internet.

[0151] "Real-time assessment of trustworthiness" refers to instantly determining the trustworthiness of content as users view it.

[0152] "Displaying a warning" means that when a user encounters unreliable content, a message is displayed to warn the user that there is a problem with the reliability of the information.

[0153] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body, and is capable of evaluating the reliability of web pages and social media content viewed by users in real time and displaying warnings. This system operates in cooperation with a server, terminals, and users.

[0154] System Overview

[0155] The server is primarily responsible for collecting, verifying, storing, learning, generating, and distributing data, while the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[0156] Server Features

[0157] 1. Content Aggregation: The server periodically retrieves trusted content from authorized sources, using public APIs and secure data transfer methods.

[0158] 2. Content verification: The retrieved content is digitally signed and authenticated, and the server verifies this to confirm authenticity.

[0159] 3. Data storage: The validated content is stored in a database, categorized according to format and content, and formatted as a dataset after removing unnecessary and duplicate data.

[0160] 4. Training a generative AI model: The shaped dataset is used to train a generative AI model, which learns patterns and relationships from large amounts of data and generates new information and answers.

[0161] 5. User request analysis: Receive and analyze user requests, using natural language processing technology to extract the intent of the question and keywords.

[0162] 6. Content search: Search for relevant content in the database based on the analysis results and extract reliable information.

[0163] 7. Answer generation: The generative AI model generates an appropriate answer based on the extracted information. The generated answer undergoes a quality check to ensure there are no problems with grammar or content.

[0164] 8. Trustworthiness Assessment: The server assesses the trustworthiness of web pages and social media content viewed by users in real time and generates a warning if the trustworthiness is low.

[0165] Device Features

[0166] The terminal primarily functions as a user interface. When the user enters a question or request, the terminal sends it to the server. When the terminal receives a response from the server, it displays the response to the user. It also displays warnings about unreliable content.

[0167] User operations

[0168] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the server's response is displayed on the terminal. In addition, the reliability of the content the user views is evaluated in real time, and a warning is displayed if the content is unreliable.

[0169] Hardware and software used

[0170] Hardware: Smartphone, Head-Mounted Display (HMD)

[0171] Frontend: React Native (Mobile App Development)

[0172] Backend: Python (Flask)

[0173] AI model: TensorFlow

[0174] Database: PostgreSQL

[0175] NLP tools: nltk, spaCy

[0176] Prompt Sentence Examples

[0177] python

[0178] Example prompt sentence:

[0179] user_input = "View the latest information about COVID-19 vaccines."

[0180] valid_sources = ["CDC", "WHO", "Ministry of Health, Labour and Welfare of Japan"]

[0181] validity_check_prompt = f"Evaluate the authenticity of the information you entered ('{user_input}') based only on data from trusted authorities ({', '.join(valid_sources)})."

[0182] This invention allows users to obtain reliable information and prevents the spread of fake news and rumours. The system's real-time evaluation and warning functions allow users to use information with peace of mind.

[0183] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0184] Step 1:

[0185] Content Collection:

[0186] The server collects content whose authenticity has been confirmed from the certification authority. Specifically, the server periodically obtains certified content using a public API or secure data transfer method. The input is content data from the certification authority, and the output is the collected content stored in the server's local storage or database.

[0187] Step 2:

[0188] Reliability verification:

[0189] The server verifies the digital signature and authentication certificate of the collected content. Specifically, it uses public key cryptography to verify the integrity of the digital signature. As a result, the input is the collected content data, and the output is content whose authenticity has been confirmed.

[0190] Step 3:

[0191] Data Retention and Classification:

[0192] The server stores the verified content in a database, classifies it according to its format and content, removes unnecessary and duplicate data, and formats it into a dataset. The input is verified content, and the output is a formatted dataset.

[0193] Step 4:

[0194] Training generative AI models:

[0195] The server uses the formatted dataset to train a generative AI model, a process that uses frameworks such as TensorFlow to learn patterns and relationships. The input is the formatted dataset, and the output is a trained generative AI model.

[0196] Step 5:

[0197] Receiving and parsing user requests:

[0198] The server receives the user request and analyzes it using natural language processing techniques (such as nltk or spaCy). The input is the user request, and the output is the analyzed question intent and keywords.

[0199] Step 6:

[0200] Search for related content:

[0201] The server searches for relevant content in the database based on the analysis results. Specifically, it uses SQL queries to extract relevant information. The input is the analyzed keywords, and the output is the related content data.

[0202] Step 7:

[0203] Answer generation:

[0204] The server generates an appropriate answer using a generative AI model based on the related content data. The generated answer undergoes quality checks (grammar and content verification) before the final answer is determined. The input is the related content data, and the output is the final answer provided to the user.

[0205] Step 8:

[0206] Reliability Rating and Warnings:

[0207] The server evaluates the trustworthiness of the web page or social media content the user is viewing in real time. If it is judged to be untrustworthy, it displays a warning on the device. The input is the content the user is currently viewing, and the output is a warning message.

[0208] 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.

[0209] This invention combines a generative AI system that uses only content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. This system aims to provide appropriate information taking into account the user's emotions, thereby achieving even higher reliability and user satisfaction. A specific embodiment of this system is described below.

[0210] System Overview

[0211] This system works in cooperation with a server, a terminal, and a user. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal sends user requests to the server and displays the server's responses to the user. The emotion engine also analyzes the emotions in the user's input text and adjusts the content and tone of the response.

[0212] Program processing

[0213] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[0214] server

[0215] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[0216] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[0217] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[0218] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the server searches for relevant content in the database and extracts reliable information.

[0219] Based on the extracted information and the results of emotion analysis, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. The answer is then adjusted based on the user's emotion recognition results, and once the final answer is determined, the server sends it to the device.

[0220] Terminal

[0221] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[0222] User

[0223] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the answer from the server is displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[0224] Specific examples

[0225] Example 1: Nutrition Question

[0226] User Input

[0227] The user inputs the question "What foods are rich in vitamin C?" into the terminal. The user is a little tired and has negative emotions.

[0228] Server Processing

[0229] The server analyzes the question and extracts the keywords "vitamin C" and "food." The emotion engine analyzes the user's emotion and recognizes that it is negative. The server then searches the database for relevant, reliable information, and the generative AI model generates an appropriate answer.

[0230] answer

[0231] The server sends an encouraging response to the device, such as, "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!", and the device displays it to the user.

[0232] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[0233] The processing flow will be explained below.

[0234] Step 1: Content Collection

[0235] server

[0236] Obtain content that is regularly verified as authentic from accredited organizations, using public APIs and secure data transfer methods.

[0237] Step 2: Verify the digital signature and authentication certificate

[0238] server

[0239] Verify the authenticity of the content by verifying the digital signature or authentication certificate attached to the content.

[0240] Step 3: Data storage and analysis

[0241] server

[0242] Content that has been verified as reliable is stored in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[0243] Step 4: Adding Metadata

[0244] server

[0245] Adding metadata to the formatted dataset improves the efficiency of data search and model training.

[0246] Step 5: Training the AI ​​model

[0247] server

[0248] Train a generative AI model using the formatted dataset in the database, then retrain the model as new data is added.

[0249] Step 6: Receiving the request

[0250] Terminal

[0251] The user enters a question or request into the terminal, which then sends the request to the server.

[0252] Step 7: Request Analysis

[0253] server

[0254] Receives requests from users and analyzes them using natural language processing technology to extract the intent of the question and keywords.

[0255] Step 8: Sentiment Analysis

[0256] server

[0257] The emotion engine analyzes the emotion of the user's input text and classifies it into positive, negative, and neutral emotion categories.

[0258] Step 9: Find related content

[0259] server

[0260] Based on the results of request analysis and sentiment analysis, relevant content is searched for in the database.

[0261] Step 10: Generate an answer

[0262] server

[0263] The generative AI model generates appropriate answers to users' questions based on the search results.

[0264] Step 11: Quality check

[0265] server

[0266] Quality check the generated answers to ensure there are no issues with grammar or content.

[0267] Step 12: Emotional Adjustment

[0268] server

[0269] Based on the user's emotion recognition results, the generated responses are adjusted to have an appropriate tone and expression.

[0270] Step 13: Submit your response

[0271] server

[0272] The final answer is sent to the device.

[0273] Step 14: Display to the User

[0274] Terminal

[0275] The answer received from the server is displayed to the user.

[0276] Specific examples

[0277] Example: Nutrition Questions

[0278] User Input

[0279] The user inputs the question "What foods are rich in vitamin C?" into the terminal.

[0280] Step 6: Receiving the request

[0281] The terminal sends a question to the server.

[0282] Step 7: Request Analysis

[0283] The server analyzes the question and extracts the keywords "vitamin C" and "food."

[0284] Step 8: Sentiment Analysis

[0285] The server's emotion engine analyzes the user's input text and recognizes that the user's emotion is negative.

[0286] Step 9: Find related content

[0287] The server retrieves relevant authoritative information from a database.

[0288] Step 10: Generate an answer

[0289] A generative AI model generates the appropriate answer.

[0290] Step 11: Quality check

[0291] The server checks the quality of the generated answers to ensure there are no problems with grammar or content.

[0292] Step 12: Emotional Adjustment

[0293] Add an encouraging tone to your responses to negative emotions.

[0294] Step 13: Submit your response

[0295] The server sends the adjusted response to the terminal.

[0296] Step 14: Display to the User

[0297] The device displays the answer to the user: "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!"

[0298] This process allows users to obtain information that is reliable and delivered in an emotionally appropriate tone, resulting in a more satisfying information experience.

[0299] Example 2

[0300] Next, a description will be given of 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."

[0301] Conventional generative AI systems may provide answers based on unreliable information, putting users at risk of receiving incorrect information. Furthermore, because they do not take user sentiment into consideration, the content and tone of their answers may be inappropriate, resulting in reduced user satisfaction. Furthermore, due to insufficient mechanisms for ensuring data quality and providing reliable information, it is difficult to prevent the spread of fake news and hoaxes.

[0302] 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.

[0303] In this invention, the server includes: means for collecting content whose reliability has been confirmed by a certification authority; means for verifying the digital signature or authentication certificate of the collected content; means for storing the content whose reliability has been confirmed in a storage device; means for training a generative AI model based on the content in the storage device; means for receiving a request from a user and analyzing the request; means for searching for related content in the storage device based on the analyzed request; an emotion engine that analyzes the user's emotions and adjusts the content and tone of a response based on the analysis results; and means for the generative AI model to generate a response based on the search results and the emotion analysis results. This makes it possible to provide highly reliable information in a tone and expression that is appropriate for the user's emotions.

[0304] "Certified Content" means information whose authenticity has been confirmed by a certification body.

[0305] A "digital signature" is a signature that is given electronically and is used to verify the authenticity and authenticity of content.

[0306] "Certificate of Authenticity" means an official certificate issued by a certification authority that indicates that content is trustworthy.

[0307] A "storage device" is a device consisting of hardware and software for storing data.

[0308] A "generative AI model" is an artificial intelligence model that uses machine learning techniques to learn patterns and relationships from large amounts of data and generate new information and answers.

[0309] A "user request" is a question or request for information that a user enters into the system.

[0310] "Natural language processing technology" is a technology for processing human language using a computer, and is used to extract the intent of a question and keywords.

[0311] The "emotion engine" is a function that analyzes the emotions in the user's input text and adjusts the content and tone of the response based on the results.

[0312] "Quality checking" refers to the process of checking generated answers for grammar and content issues.

[0313] "Public API" means an application program interface that is made publicly available for use by external developers and applications.

[0314] A "secure data transfer method" is a data transfer method that employs security measures such as encryption to prevent data eavesdropping or tampering.

[0315] This invention combines a generative AI system that uses content whose authenticity has been confirmed by a certification organization with an emotion engine that recognizes user emotions. The system is designed to operate in cooperation with the server, terminal, and user.

[0316] System configuration

[0317] server

[0318] The server is responsible for collecting, storing, analyzing, generating, and distributing data. Specifically, it operates in the following steps:

[0319] 1. Content collection: The server periodically obtains content whose authenticity has been verified from a certification authority. The server obtains data using public APIs and secure data transfer methods. For example, "https: / / api.certifiedcontent.org / data" is used as the public API.

[0320] 2. Authenticity verification: Verify the authenticity of the retrieved content by verifying its digital signature or authentication certificate.

[0321] 3. Data storage and formatting: Once content is verified as reliable, it is stored on a storage device, categorized by format and content, and redundant and duplicate data is removed. The organized dataset is then given metadata to enable efficient searching and learning.

[0322] 4. Training a generative AI model: A generative AI model is trained using the shaped dataset, using a framework such as TensorFlow or PyTorch, and is retrained every time new data is added.

[0323] Terminal

[0324] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays it to the user. The response is adjusted by the emotion engine and is delivered in a tone and expression that matches the user's emotions.

[0325] User

[0326] Users access the system through a terminal. They input questions or requests for information into the terminal, and the terminal sends the request to the server. The server's response is displayed on the terminal. The user can receive responses tailored by the emotion engine, resulting in a more satisfying experience.

[0327] Specific examples

[0328] Example 1: Nutrition Question

[0329] User Input

[0330] The user inputs the question "What foods are rich in vitamin C?" into the terminal. It is assumed that the user is slightly tired and in a negative emotional state.

[0331] Server Processing

[0332] The server receives the question and uses natural language processing technology to extract keywords such as "vitamin C" and "food." The emotion engine analyzes the user's emotions and recognizes negative emotions. The server then searches the database for relevant and reliable information, and the generative AI model generates an appropriate answer.

[0333] answer

[0334] The server generates a response such as the following and sends it to the device: "Foods that are rich in vitamin C include oranges, kiwis, and bell peppers. Eating these foods will help you stay healthy!" The response has an encouraging tone and is displayed on the device to the user.

[0335] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[0336] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0337] Step 1:

[0338] The server collects content whose authenticity has been confirmed from a certification authority. The input is data obtained through a public API or secure data transfer method. Specifically, the server uses the public API to call "https: / / api.certifiedcontent.org / data" and receives data in JSON format. The output is the raw data before its authenticity is confirmed.

[0339] Step 2:

[0340] The server verifies the digital signatures and authentication certificates of the collected content. The input is the content collected in step 1. Specifically, it verifies its authenticity using an algorithm that verifies the digital signature. The output is the data whose authenticity has been confirmed.

[0341] Step 3:

[0342] The server saves the content whose authenticity has been confirmed in the storage device. The input is the data whose authenticity has been confirmed in step 2. Specifically, it executes the "save" command to store the data in the database. The output is the trusted information as the data stored in the database.

[0343] Step 4:

[0344] The server analyzes the stored content and classifies it according to format and content. The input is the data in the database stored in step 3. It uses queries to classify the data into categories such as "nutrition," "health," and "emotions," and then refines the dataset by removing unnecessary and duplicate data. The output is the refined dataset.

[0345] Step 5:

[0346] The server trains the generative AI model using the formatted dataset. The input is the formatted dataset from step 4. Specifically, it runs a Python script and trains the model using TensorFlow or PyTorch. If necessary, it also retrains the model when new data is added. The output is a trained generative AI model.

[0347] Step 6:

[0348] The user inputs a question or request through the terminal. The input is text data representing the user's question or request. The terminal sends this request to the server. Specific operations include text input and transmission functions through the terminal's UI. The output is the request data sent to the server.

[0349] Step 7:

[0350] The server receives a request from the user and analyzes it. The input is the request data sent in step 6. Specifically, it uses a natural language processing library such as "NLTK" or "spaCy" to extract the intent and keywords of the question. The output is the analyzed intent and keyword information of the question.

[0351] Step 8:

[0352] The server runs an emotion engine that analyzes the emotion of the user's input text. The input is the question intent and keyword information extracted in step 7. The emotion analysis module is used to classify the emotion into positive, negative, and neutral categories. The output is data indicating the user's emotional state.

[0353] Step 9:

[0354] The server searches for relevant content in the database based on the analyzed request and emotional information. The input is the emotional state and question intent / keyword information obtained in step 8. It executes an SQL query to retrieve relevant information. The output is reliable relevant information.

[0355] Step 10:

[0356] The server uses the generative AI model to generate an answer based on the search results and sentiment analysis results. The input is the relevant information and emotional state obtained in step 9. The generative AI model inputs a prompt sentence to generate an appropriate answer. As a specific example, a "positive prompt sentence" is used to generate an answer with a positive tone. The output is the generated answer.

[0357] Step 11:

[0358] The server quality checks the generated answer to ensure there are no problems with grammar or content. The input is the answer generated in step 10. A quality check module is run to check the answer for accuracy and appropriateness. The output is the final answer that has passed the quality check.

[0359] Step 12:

[0360] The server sends the final answer to the device, which displays the answer to the user. The input is the final answer that passed the quality check in step 11. The server uses the Send API to send the answer to the device and updates the device's UI elements to display it to the user. The output is the answer displayed on the device in a form that the user can see.

[0361] (Application example 2)

[0362] Next, a description will be given of Application 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."

[0363] Conventional generative AI systems have not adequately considered the reliability of content and user sentiment, resulting in a poor user experience and inaccurate information provided. Furthermore, the use of unreliable information can lead to problems such as the provision of inaccurate information and the spread of fake news.

[0364] 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.

[0365] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the content whose reliability has been confirmed in a database, means for training a generative AI model based on the content in the database, means for recognizing and analyzing a user's emotions, means for adjusting the generated answers based on the analyzed emotions, means for quality checking the generated answers and checking for problems with grammar and content, and means for using a public API and secure data transfer means. This makes it possible to provide highly reliable information and optimal information tailored to the user's emotions.

[0366] A "generative AI model" is an algorithm that learns from collected data and generates new information and answers.

[0367] A "certification body" is an organization whose role is to verify and certify the reliability and accuracy of content.

[0368] "Confirmed reliable content" is information whose accuracy and integrity have been confirmed by a certification body.

[0369] A "digital signature" is an electronic means of verifying the origin of content or data and ensuring that it has not been tampered with.

[0370] "Certificate of Accreditation" means official evidence of an accreditation body's confirmation of trustworthiness.

[0371] An "emotion engine" is software that analyzes and classifies emotions from user input text and voice.

[0372] A "database" is a system that organizes and stores collected content and data.

[0373] "Quality check" is the process of checking the generated content for grammar and content appropriateness.

[0374] A "public API" is an interface designed to allow external access to specific functions or data.

[0375] "Secure data transfer methods" are technologies and protocols that prevent unauthorized access or tampering during data transmission.

[0376] A "prompt" is text that instructs a generative AI model on the format and content of input data.

[0377] This invention is a system that combines a generative AI system that uses content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. Specific embodiments of this system are described below.

[0378] Setup and Basic Operation

[0379] The main components of the system are the server, the terminals, and the users, each of which plays a specific role.

[0380] server

[0381] The server is responsible for collecting, validating, storing, learning, generating, and distributing data. Specifically, it works as follows:

[0382] 1. Content collection and verification: The server collects content that has been verified as trustworthy from a trusted authority, using public APIs and secure data transfer methods. The collected content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[0383] 2. Storage and classification in a database: Once the content has been verified as reliable, it is stored in a database. The stored content is analyzed and classified according to its format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[0384] 3. Training the generative AI model: The generative AI model is trained using the shaped dataset. The generative AI model learns patterns and relationships from large amounts of data and has the ability to generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[0385] 4. Request analysis and answer generation: Upon receiving a request from a user, the server analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the generative AI model generates an answer, and a quality check is performed to determine the final answer.

[0386] Terminal

[0387] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[0388] User

[0389] Users access the system through a terminal. They input questions or requests for information into the terminal, and the answers from the server are displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[0390] Specific examples

[0391] Suppose a user puts on smart glasses and inputs, "I'd like to know how to relax after a long day at work." This input is sent to the server via the device. The server analyzes the input, and its emotion engine classifies the user's emotional state as "fatigue." As a result, the server searches for appropriate product information (e.g., aroma diffuser or relaxation massager) from a trusted product database, and generates the following suggestion using a generative AI model: "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or relaxation massager. This aroma diffuser, in particular, has a lavender scent that is expected to promote sound sleep."

[0392] Prompt Sentence Examples

[0393] 1. Prompt to analyze emotions from user input:

[0394] User input text: "I want to know how to relax after a long day at work." Analyze the user's emotional state from this text.

[0395] 2. Prompt to generate sentiment-based product suggestions:

[0396] The user's emotional state has been analyzed as "fatigue." Please generate a sentence to explain to the user about a reliable aroma diffuser or relaxing massager.

[0397] In this way, users receive reliable information in a manner appropriate to their emotional state.

[0398] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0399] Step 1:

[0400] The device receives a user's question or request. The device receives input data by inputting a specific question or request in voice or text format. An example input might be a request such as, "I'd like to know how to relax after a long day at work." The input data is then sent directly to the next step.

[0401] Step 2:

[0402] The device sends the received user question or request to the server, using an Internet connection to transmit the user's input data to the server, which then receives the requested data.

[0403] Step 3:

[0404] The server analyzes the received request. Natural language processing technology is used to extract the intent of the question and key keywords from the text. For example, keywords such as "how to relax" and "long hours at work" may be extracted. Based on the results of this analysis, the system proceeds to the next processing step.

[0405] Step 4:

[0406] The server uses the emotion engine to analyze the emotion of the user's input text. For example, if the user's input is "I want to know how to relax after a long day of work," the emotion engine will classify the user's emotional state as "fatigue" from this text. The analysis result is saved as an emotion category.

[0407] Step 5:

[0408] The server searches for relevant information from a trusted content database based on the analyzed emotional state and key keywords. For example, it searches for content related to "how to relax" or "fatigue" and retrieves the results. These results are then used.

[0409] Step 6:

[0410] The server generates an answer based on the search results using a generative AI model. The generative AI model uses patterns and relationships learned from large amounts of data to generate an appropriate answer. For example, a sentence such as "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or a relaxing massage machine" is generated. This generated sentence is then used.

[0411] Step 7:

[0412] The server then performs a quality check on the generated answer, checking for grammar and content issues and making corrections as necessary. Once the answer passes this check, it is used.

[0413] Step 8:

[0414] The server then adjusts the generated answer based on the results of the emotion analysis. To match the tired emotion, the server may change the tone to something gentler and more relaxing. The final answer is then sent in the next step.

[0415] Step 9:

[0416] The server sends the final answer to the device, using an Internet connection to transmit the answer data to the device, which can then receive the data from the server.

[0417] Step 10:

[0418] The device then displays the received response to the user. The response may be in the form of text or voice, and is delivered in an appropriate tone and expression. For example, a message such as "Thank you for your long hours at work. To help you relax, we recommend using an aroma diffuser or a relaxing massager" may be displayed.

[0419] 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.

[0420] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[0421] 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.

[0422] [Second embodiment]

[0423] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0424] 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.

[0425] 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).

[0426] 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.

[0427] 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.

[0428] 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).

[0429] 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.

[0430] 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.

[0431] 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.

[0432] 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.

[0433] 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.

[0434] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0435] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body. The purpose of this system is to provide reliable information and prevent the spread of fake news and hoaxes. A specific embodiment of this system is described below.

[0436] System Overview

[0437] This system involves a server, a terminal, and a user working together. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[0438] Program processing

[0439] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[0440] server

[0441] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[0442] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[0443] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[0444] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. Based on this, the server searches for relevant content in the database and extracts reliable information.

[0445] Based on the extracted information, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. Once the final answer is determined, the server sends it to the device.

[0446] Terminal

[0447] The terminal acts as the user's interface: the user enters a question or request, which the terminal sends to the server, and upon receiving the answer from the server, the terminal displays the answer to the user.

[0448] User

[0449] Users access the system through terminals, which they type into the terminal questions or requests for information, which then transmit the requests to the server, and the server's responses are displayed on the terminal.

[0450] Specific examples

[0451] Example 1: Medical information question

[0452] User Input

[0453] The user inputs the question "How can I relieve cold symptoms?" into the terminal.

[0454] Server Processing

[0455] The server analyzes the question and searches a database for relevant and reliable medical information. Based on information from certified medical institutions, a generative AI model generates an appropriate answer.

[0456] answer

[0457] The server sends an answer such as, "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays the answer to the user.

[0458] This invention allows users to obtain reliable information and prevents the spread of fake news and rumors.

[0459] The processing flow will be explained below.

[0460] Step 1: Please give me some.

[0461] Example 1

[0462] Next, a description will be given of 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."

[0463] Today, the internet is flooded with unreliable information and fake news, making it difficult for users to quickly obtain reliable information. This problem is particularly severe in specialized fields such as medicine and law, where false information can have a significant impact. Furthermore, conventional systems do not fully establish procedures for verifying reliability, formatting data, and checking the quality of generated answers, meaning the reliability of the information provided to users cannot be guaranteed. To address these issues, a generative AI system is needed that can provide reliable information and prevent the spread of fake news and rumours.

[0464] 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.

[0465] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the verified content in a database, means for analyzing the stored content, classifying it according to format and content, deleting unnecessary data and duplicate data, and formatting it into a dataset, means for training a generative AI model based on the formatted dataset in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed request, means for the generative AI model to generate an answer based on the search results, and means for quality checking the generated answer to check for problems with grammar and content. This allows users to obtain reliable information quickly and accurately.

[0466] A "certification body" is an organization that verifies and certifies the accuracy and reliability of specific content or information with the aim of providing reliable information.

[0467] A "digital signature" is an electronic means of authentication that uses cryptographic technology to verify the identity of the sender of electronic information and whether the information has been tampered with.

[0468] A "certificate of accreditation" is a paper or digital certificate issued by an accreditation body to officially certify the authenticity or accuracy of particular content.

[0469] A "database" is a system for efficiently and systematically storing and managing information and data.

[0470] A "generative AI model" is an artificial intelligence algorithm that learns patterns and relationships from large amounts of data and generates new information and answers.

[0471] A "user request" is a question or request that a user sends to the system for specific information.

[0472] "Natural language processing" is the technology that enables computers to understand, analyze, and generate human language.

[0473] "Quality check" is an inspection process to check whether there are any problems with the grammar or content of the generated answer and to provide it to the user in the most optimal form.

[0474] A "public API" is a public application programming interface that provides an interface for external developers to access specified functionality and data.

[0475] A "secure data transfer method" is a data transfer method that uses encryption and authentication to prevent information leaks and unauthorized access when sending and receiving data.

[0476] This invention is a generative AI system designed to provide reliable information and prevent the spread of fake news and hoaxes. This system operates in cooperation with a server, terminals, and users.

[0477] The server performs processing using the following hardware and software:

[0478] Hardware: High-performance processor, memory, storage, network interface

[0479] Software: Public APIs, secure data transfer methods (e.g., TLS), digital signature verification tools (e.g., OpenSSL), databases (e.g., MySQL), analytics scripts (e.g., Python and the pandas library), generative AI models (e.g., TensorFlow), and natural language processing tools (e.g., Google Cloud Natural Language API and Grammarly API).

[0480] 1. Content Collection

[0481] The server runs a scheduled task to collect trusted content from trusted organizations and sends HTTP GET requests to designated API endpoints. For example, to retrieve medical information, the server retrieves data from the API of a trusted medical institution.

[0482] 2. Reliability check

[0483] The server verifies the digital signature and authentication certificate of the retrieved content. This process uses "OpenSSL" to verify whether the content signature is correct, for example, by using the "openssl verify" command.

[0484] 3. Saving to the database and formatting

[0485] Once the content is verified as reliable, it is stored in a MySQL database. Python scripts are then run to parse the data, classify it by format and content, and clean it up by removing unnecessary and duplicate data, allowing for efficient searching and learning.

[0486] 4. Training the generative AI model

[0487] The resulting dataset is used to train a generative AI model using the TensorFlow library, and each time new data is added to the system, the model is retrained to improve its accuracy.

[0488] 5. User Request Analysis

[0489] The server receives questions sent by users from their devices and analyzes them using natural language processing technology. The server uses the Google Cloud Natural Language API to extract intent and keywords.

[0490] 6. Extraction of reliable information and generation of answers

[0491] The server searches for relevant content in the database and generates an appropriate answer using a generative AI model. This generative AI model generates an answer based on a "prompt sentence." For example, if a user enters "How can I relieve cold symptoms?", the prompt sentence is set to "Provide appropriate medical information based on the user's question."

[0492] 7. Quality check of answers

[0493] The generated answers are quality checked using the Grammarly API to ensure there are no problems with grammar or content.

[0494] 8. Submitting and Viewing Your Answers

[0495] The final answer is sent from the server to the device as an HTTP response, and the device displays it to the user, who can check the answer through the device's interface.

[0496] Specific examples

[0497] Example: Questions about medical information

[0498] User input:

[0499] The user types into the terminal, "How can I relieve my cold symptoms?"

[0500] Server Action:

[0501] The server analyzes the question using the Google Cloud Natural Language API and extracts keywords (e.g., "cold," "symptoms," "relieve").

[0502] The server searches the database for relevant and reliable medical information.

[0503] The generative AI model uses the prompt "How to relieve cold symptoms" to generate an answer.

[0504] The generated answers are quality checked using the Grammarly API.

[0505] answer:

[0506] The server sends the answer "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays it to the user.

[0507] This invention allows users to obtain reliable information quickly and accurately, and eliminates the influence of fake news and hoaxes.

[0508] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0509] Step 1: Content Collection

[0510] The server collects trusted content from a certification authority. As input, it uses the public API endpoint URL. Specifically, the server runs a scheduled task (e.g., a cron job) to send an HTTP GET request. The output is the retrieved content data.

[0511] Step 2: Authenticity check

[0512] The server verifies the digital signature and authentication certificate of the retrieved content. It uses the retrieved content data and its digital signature as input. Specifically, the server uses the "OpenSSL" library to verify the digital signature. The output is the content, which has been confirmed to be authentic.

[0513] Step 3: Saving to the database and formatting

[0514] The server stores the verified content in a database and formats it. The verified content data is used as input. Specifically, the server stores the data in a MySQL database, runs a Python script to analyze the data, and classifies it according to format and content. It then removes unnecessary and duplicate data and formats it into a dataset. The output is a formatted dataset.

[0515] Step 4: Training the generative AI model

[0516] The server trains the generative AI model using the formatted dataset. The formatted dataset is used as input. Specifically, the server retrains the generative AI model using the TensorFlow library. The output is a trained generative AI model.

[0517] Step 5: Receiving a User Request

[0518] The terminal receives requests from the user. As input, it uses questions or requests that the user types into the terminal. Specifically, the terminal receives input through a user interface and sends it to the server. The output is the user request sent to the server.

[0519] Step 6: Parsing the user request

[0520] The server receives the user request and analyzes it using natural language processing technology. The user request is used as input. Specifically, the server uses the Google Cloud Natural Language API to extract the intent and keywords of the request. The output is the analyzed intent and keywords.

[0521] Step 7: Extract reliable information

[0522] The server searches for relevant content in a database based on the parsed intent and keywords. It uses the parsed intent, keywords, and database as input. Specifically, the server executes SQL queries to extract relevant information. The output is the relevant content.

[0523] Step 8: Answer Generation

[0524] The server uses a generative AI model to generate an appropriate answer based on related content. The server uses related content and the generative AI model as input. Specifically, it inputs a "prompt sentence" into the generative AI model to generate an answer. The output is the generated answer. For example, the prompt sentence could be "Provide appropriate medical information based on the user's question."

[0525] Step 9: Check the quality of your answers

[0526] The server checks the quality of the generated answer. It uses the generated answer as input. Specifically, the server checks the grammar and content using the Grammarly API. The output is a quality-checked answer.

[0527] Step 10: Submit and view your responses

[0528] The server sends the quality-checked answer to the terminal, which displays it to the user. The quality-checked answer is used as input. Specifically, the server generates an HTTP response and sends it to the terminal. The terminal analyzes the response data and displays it in its user interface. The output is the answer displayed to the user.

[0529] (Application example 1)

[0530] Next, a description will be given of Application 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."

[0531] In today's information society, the spread of fake news and inaccurate information has become a serious problem. This increases the risk that users will make incorrect decisions based on unreliable information. In particular, unreliable information can easily spread on web pages and social media, so there is a need for a mechanism to quickly detect this and notify users.

[0532] 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.

[0533] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and authentication certificates of the collected content, means for storing the confirmed reliability content in a database, means for training a generative AI model based on the content in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed requests, means for the generative AI model to generate an answer based on the search results, means for evaluating the reliability of content on web pages and social media that users are viewing in real time, and means for displaying warnings for unreliable content. This allows users to make decisions based on reliable information and prevents the spread of fake news and hoaxes.

[0534] An "accreditation body" is an organization or group whose role is to officially verify and certify the reliability of information.

[0535] "Authenticated content" means information or data that has been formally verified and certified for accuracy and authenticity by an accredited organization.

[0536] A "digital signature" is an electronic certificate that is given to ensure the authenticity and integrity of content.

[0537] A "certificate of authenticity" is an official certification that particular content comes from a trusted source.

[0538] A "database" is a collection of information that stores collected content in an organized manner and can be efficiently searched and used.

[0539] A "generative AI model" is an artificial intelligence model that has the ability to learn patterns and relationships from large amounts of data and generate new information and answers.

[0540] A "request" is an inquiry or request that a user sends to the system for information.

[0541] "Analysis" is the process of extracting the intent and keywords of the user's request and understanding their meaning.

[0542] "Related content" is information or data that is relevant to the user's request.

[0543] An "answer" is appropriate information or a solution generated based on a user request.

[0544] A "web page" is a collection of information or data that can be viewed on the Internet.

[0545] "Social media" refers to online platforms that allow users to share information and communicate over the internet.

[0546] "Real-time assessment of trustworthiness" refers to instantly determining the trustworthiness of content as users view it.

[0547] "Displaying a warning" means that when a user encounters unreliable content, a message is displayed to warn the user that there is a problem with the reliability of the information.

[0548] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body, and is capable of evaluating the reliability of web pages and social media content viewed by users in real time and displaying warnings. This system operates in cooperation with a server, terminals, and users.

[0549] System Overview

[0550] The server is primarily responsible for collecting, verifying, storing, learning, generating, and distributing data, while the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[0551] Server Features

[0552] 1. Content Aggregation: The server periodically retrieves trusted content from authorized sources, using public APIs and secure data transfer methods.

[0553] 2. Content verification: The retrieved content is digitally signed and authenticated, and the server verifies this to confirm authenticity.

[0554] 3. Data storage: The validated content is stored in a database, categorized according to format and content, and formatted as a dataset after removing unnecessary and duplicate data.

[0555] 4. Training a generative AI model: The shaped dataset is used to train a generative AI model, which learns patterns and relationships from large amounts of data and generates new information and answers.

[0556] 5. User request analysis: Receive and analyze user requests, using natural language processing technology to extract the intent of the question and keywords.

[0557] 6. Content search: Search for relevant content in the database based on the analysis results and extract reliable information.

[0558] 7. Answer generation: The generative AI model generates an appropriate answer based on the extracted information. The generated answer undergoes a quality check to ensure there are no problems with grammar or content.

[0559] 8. Trustworthiness Assessment: The server assesses the trustworthiness of web pages and social media content viewed by users in real time and generates a warning if the trustworthiness is low.

[0560] Device Features

[0561] The terminal primarily functions as a user interface. When the user enters a question or request, the terminal sends it to the server. When the terminal receives a response from the server, it displays the response to the user. It also displays warnings about unreliable content.

[0562] User operations

[0563] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the server's response is displayed on the terminal. In addition, the reliability of the content the user views is evaluated in real time, and a warning is displayed if the content is unreliable.

[0564] Hardware and software used

[0565] Hardware: Smartphone, Head-Mounted Display (HMD)

[0566] Frontend: React Native (Mobile App Development)

[0567] Backend: Python (Flask)

[0568] AI model: TensorFlow

[0569] Database: PostgreSQL

[0570] NLP tools: nltk, spaCy

[0571] Prompt Sentence Examples

[0572] python

[0573] Example prompt sentence:

[0574] user_input = "View the latest information about COVID-19 vaccines."

[0575] valid_sources = ["CDC", "WHO", "Ministry of Health, Labour and Welfare of Japan"]

[0576] validity_check_prompt = f"Evaluate the authenticity of the information you entered ('{user_input}') based only on data from trusted authorities ({', '.join(valid_sources)})."

[0577] This invention allows users to obtain reliable information and prevents the spread of fake news and rumours. The system's real-time evaluation and warning functions allow users to use information with peace of mind.

[0578] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0579] Step 1:

[0580] Content Collection:

[0581] The server collects content whose authenticity has been confirmed from the certification authority. Specifically, the server periodically obtains certified content using a public API or secure data transfer method. The input is content data from the certification authority, and the output is the collected content stored in the server's local storage or database.

[0582] Step 2:

[0583] Reliability verification:

[0584] The server verifies the digital signature and authentication certificate of the collected content. Specifically, it uses public key cryptography to verify the integrity of the digital signature. As a result, the input is the collected content data, and the output is content whose authenticity has been confirmed.

[0585] Step 3:

[0586] Data Retention and Classification:

[0587] The server stores the verified content in a database, classifies it according to its format and content, removes unnecessary and duplicate data, and formats it into a dataset. The input is verified content, and the output is a formatted dataset.

[0588] Step 4:

[0589] Training generative AI models:

[0590] The server uses the formatted dataset to train a generative AI model, a process that uses frameworks such as TensorFlow to learn patterns and relationships. The input is the formatted dataset, and the output is a trained generative AI model.

[0591] Step 5:

[0592] Receiving and parsing user requests:

[0593] The server receives the user request and analyzes it using natural language processing techniques (such as nltk or spaCy). The input is the user request, and the output is the analyzed question intent and keywords.

[0594] Step 6:

[0595] Search for related content:

[0596] The server searches for relevant content in the database based on the analysis results. Specifically, it uses SQL queries to extract relevant information. The input is the analyzed keywords, and the output is the related content data.

[0597] Step 7:

[0598] Answer generation:

[0599] The server generates an appropriate answer using a generative AI model based on the related content data. The generated answer undergoes quality checks (grammar and content verification) before the final answer is determined. The input is the related content data, and the output is the final answer provided to the user.

[0600] Step 8:

[0601] Reliability Rating and Warnings:

[0602] The server evaluates the trustworthiness of the web page or social media content the user is viewing in real time. If it is judged to be untrustworthy, it displays a warning on the device. The input is the content the user is currently viewing, and the output is a warning message.

[0603] 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.

[0604] This invention combines a generative AI system that uses only content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. This system aims to provide appropriate information taking into account the user's emotions, thereby achieving even higher reliability and user satisfaction. A specific embodiment of this system is described below.

[0605] System Overview

[0606] This system works in cooperation with a server, a terminal, and a user. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal sends user requests to the server and displays the server's responses to the user. The emotion engine also analyzes the emotions in the user's input text and adjusts the content and tone of the response.

[0607] Program processing

[0608] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[0609] server

[0610] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[0611] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[0612] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[0613] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the server searches for relevant content in the database and extracts reliable information.

[0614] Based on the extracted information and the results of emotion analysis, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. The answer is then adjusted based on the user's emotion recognition results, and once the final answer is determined, the server sends it to the device.

[0615] Terminal

[0616] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[0617] User

[0618] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the answer from the server is displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[0619] Specific examples

[0620] Example 1: Nutrition Question

[0621] User Input

[0622] The user inputs the question "What foods are rich in vitamin C?" into the terminal. The user is a little tired and has negative emotions.

[0623] Server Processing

[0624] The server analyzes the question and extracts the keywords "vitamin C" and "food." The emotion engine analyzes the user's emotion and recognizes that it is negative. The server then searches the database for relevant, reliable information, and the generative AI model generates an appropriate answer.

[0625] answer

[0626] The server sends an encouraging response to the device, such as, "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!", and the device displays it to the user.

[0627] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[0628] The processing flow will be explained below.

[0629] Step 1: Content Collection

[0630] server

[0631] Obtain content that is regularly verified as authentic from accredited organizations, using public APIs and secure data transfer methods.

[0632] Step 2: Verify the digital signature and authentication certificate

[0633] server

[0634] Verify the authenticity of the content by verifying the digital signature or authentication certificate attached to the content.

[0635] Step 3: Data storage and analysis

[0636] server

[0637] Content that has been verified as reliable is stored in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[0638] Step 4: Adding Metadata

[0639] server

[0640] Adding metadata to the formatted dataset improves the efficiency of data search and model training.

[0641] Step 5: Training the AI ​​model

[0642] server

[0643] Train a generative AI model using the formatted dataset in the database, then retrain the model as new data is added.

[0644] Step 6: Receiving the request

[0645] Terminal

[0646] The user enters a question or request into the terminal, which then sends the request to the server.

[0647] Step 7: Request Analysis

[0648] server

[0649] Receives requests from users and analyzes them using natural language processing technology to extract the intent of the question and keywords.

[0650] Step 8: Sentiment Analysis

[0651] server

[0652] The emotion engine analyzes the emotion of the user's input text and classifies it into positive, negative, and neutral emotion categories.

[0653] Step 9: Find related content

[0654] server

[0655] Based on the results of request analysis and sentiment analysis, relevant content is searched for in the database.

[0656] Step 10: Generate an answer

[0657] server

[0658] The generative AI model generates appropriate answers to users' questions based on the search results.

[0659] Step 11: Quality check

[0660] server

[0661] Quality check the generated answers to ensure there are no issues with grammar or content.

[0662] Step 12: Emotional Adjustment

[0663] server

[0664] Based on the user's emotion recognition results, the generated responses are adjusted to have an appropriate tone and expression.

[0665] Step 13: Submit your response

[0666] server

[0667] The final answer is sent to the device.

[0668] Step 14: Display to the User

[0669] Terminal

[0670] The answer received from the server is displayed to the user.

[0671] Specific examples

[0672] Example: Nutrition Questions

[0673] User Input

[0674] The user inputs the question "What foods are rich in vitamin C?" into the terminal.

[0675] Step 6: Receiving the request

[0676] The terminal sends a question to the server.

[0677] Step 7: Request Analysis

[0678] The server analyzes the question and extracts the keywords "vitamin C" and "food."

[0679] Step 8: Sentiment Analysis

[0680] The server's emotion engine analyzes the user's input text and recognizes that the user's emotion is negative.

[0681] Step 9: Find related content

[0682] The server retrieves relevant authoritative information from a database.

[0683] Step 10: Generate an answer

[0684] A generative AI model generates the appropriate answer.

[0685] Step 11: Quality check

[0686] The server checks the quality of the generated answers to ensure there are no problems with grammar or content.

[0687] Step 12: Emotional Adjustment

[0688] Add an encouraging tone to your responses to negative emotions.

[0689] Step 13: Submit your response

[0690] The server sends the adjusted response to the terminal.

[0691] Step 14: Display to the User

[0692] The device displays the answer to the user: "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!"

[0693] This process allows users to obtain information that is reliable and delivered in an emotionally appropriate tone, resulting in a more satisfying information experience.

[0694] Example 2

[0695] Next, a description will be given of 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."

[0696] Conventional generative AI systems may provide answers based on unreliable information, putting users at risk of receiving incorrect information. Furthermore, because they do not take user sentiment into consideration, the content and tone of their answers may be inappropriate, resulting in reduced user satisfaction. Furthermore, due to insufficient mechanisms for ensuring data quality and providing reliable information, it is difficult to prevent the spread of fake news and hoaxes.

[0697] 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.

[0698] In this invention, the server includes: means for collecting content whose reliability has been confirmed by a certification authority; means for verifying the digital signature or authentication certificate of the collected content; means for storing the content whose reliability has been confirmed in a storage device; means for training a generative AI model based on the content in the storage device; means for receiving a request from a user and analyzing the request; means for searching for related content in the storage device based on the analyzed request; an emotion engine that analyzes the user's emotions and adjusts the content and tone of a response based on the analysis results; and means for the generative AI model to generate a response based on the search results and the emotion analysis results. This makes it possible to provide highly reliable information in a tone and expression that is appropriate for the user's emotions.

[0699] "Certified Content" means information whose authenticity has been confirmed by a certification body.

[0700] A "digital signature" is a signature that is given electronically and is used to verify the authenticity and authenticity of content.

[0701] "Certificate of Authenticity" means an official certificate issued by a certification authority that indicates that content is trustworthy.

[0702] A "storage device" is a device consisting of hardware and software for storing data.

[0703] A "generative AI model" is an artificial intelligence model that uses machine learning techniques to learn patterns and relationships from large amounts of data and generate new information and answers.

[0704] A "user request" is a question or request for information that a user enters into the system.

[0705] "Natural language processing technology" is a technology for processing human language using a computer, and is used to extract the intent of a question and keywords.

[0706] The "emotion engine" is a function that analyzes the emotions in the user's input text and adjusts the content and tone of the response based on the results.

[0707] "Quality checking" refers to the process of checking generated answers for grammar and content issues.

[0708] "Public API" means an application program interface that is made publicly available for use by external developers and applications.

[0709] A "secure data transfer method" is a data transfer method that employs security measures such as encryption to prevent data eavesdropping or tampering.

[0710] This invention combines a generative AI system that uses content whose authenticity has been confirmed by a certification organization with an emotion engine that recognizes user emotions. The system is designed to operate in cooperation with the server, terminal, and user.

[0711] System configuration

[0712] server

[0713] The server is responsible for collecting, storing, analyzing, generating, and distributing data. Specifically, it operates in the following steps:

[0714] 1. Content collection: The server periodically obtains content whose authenticity has been verified from a certification authority. The server obtains data using public APIs and secure data transfer methods. For example, "https: / / api.certifiedcontent.org / data" is used as the public API.

[0715] 2. Authenticity verification: Verify the authenticity of the retrieved content by verifying its digital signature or authentication certificate.

[0716] 3. Data storage and formatting: Once content is verified as reliable, it is stored on a storage device, categorized by format and content, and redundant and duplicate data is removed. The organized dataset is then given metadata to enable efficient searching and learning.

[0717] 4. Training a generative AI model: A generative AI model is trained using the shaped dataset, using a framework such as TensorFlow or PyTorch, and is retrained every time new data is added.

[0718] Terminal

[0719] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays it to the user. The response is adjusted by the emotion engine and is delivered in a tone and expression that matches the user's emotions.

[0720] User

[0721] Users access the system through a terminal. They input questions or requests for information into the terminal, and the terminal sends the request to the server. The server's response is displayed on the terminal. The user can receive responses tailored by the emotion engine, resulting in a more satisfying experience.

[0722] Specific examples

[0723] Example 1: Nutrition Question

[0724] User Input

[0725] The user inputs the question "What foods are rich in vitamin C?" into the terminal. It is assumed that the user is slightly tired and in a negative emotional state.

[0726] Server Processing

[0727] The server receives the question and uses natural language processing technology to extract keywords such as "vitamin C" and "food." The emotion engine analyzes the user's emotions and recognizes negative emotions. The server then searches the database for relevant and reliable information, and the generative AI model generates an appropriate answer.

[0728] answer

[0729] The server generates a response such as the following and sends it to the device: "Foods that are rich in vitamin C include oranges, kiwis, and bell peppers. Eating these foods will help you stay healthy!" The response has an encouraging tone and is displayed on the device to the user.

[0730] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[0731] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0732] Step 1:

[0733] The server collects content whose authenticity has been confirmed from a certification authority. The input is data obtained through a public API or secure data transfer method. Specifically, the server uses the public API to call "https: / / api.certifiedcontent.org / data" and receives data in JSON format. The output is the raw data before its authenticity is confirmed.

[0734] Step 2:

[0735] The server verifies the digital signatures and authentication certificates of the collected content. The input is the content collected in step 1. Specifically, it verifies its authenticity using an algorithm that verifies the digital signature. The output is the data whose authenticity has been confirmed.

[0736] Step 3:

[0737] The server saves the content whose authenticity has been confirmed in the storage device. The input is the data whose authenticity has been confirmed in step 2. Specifically, it executes the "save" command to store the data in the database. The output is the trusted information as the data stored in the database.

[0738] Step 4:

[0739] The server analyzes the stored content and classifies it according to format and content. The input is the data in the database stored in step 3. It uses queries to classify the data into categories such as "nutrition," "health," and "emotions," and then refines the dataset by removing unnecessary and duplicate data. The output is the refined dataset.

[0740] Step 5:

[0741] The server trains the generative AI model using the formatted dataset. The input is the formatted dataset from step 4. Specifically, it runs a Python script and trains the model using TensorFlow or PyTorch. If necessary, it also retrains the model when new data is added. The output is a trained generative AI model.

[0742] Step 6:

[0743] The user inputs a question or request through the terminal. The input is text data representing the user's question or request. The terminal sends this request to the server. Specific operations include text input and transmission functions through the terminal's UI. The output is the request data sent to the server.

[0744] Step 7:

[0745] The server receives a request from the user and analyzes it. The input is the request data sent in step 6. Specifically, it uses a natural language processing library such as "NLTK" or "spaCy" to extract the intent and keywords of the question. The output is the analyzed intent and keyword information of the question.

[0746] Step 8:

[0747] The server runs an emotion engine that analyzes the emotion of the user's input text. The input is the question intent and keyword information extracted in step 7. The emotion analysis module is used to classify the emotion into positive, negative, and neutral categories. The output is data indicating the user's emotional state.

[0748] Step 9:

[0749] The server searches for relevant content in the database based on the analyzed request and emotional information. The input is the emotional state and question intent / keyword information obtained in step 8. It executes an SQL query to retrieve relevant information. The output is reliable relevant information.

[0750] Step 10:

[0751] The server uses the generative AI model to generate an answer based on the search results and sentiment analysis results. The input is the relevant information and emotional state obtained in step 9. The generative AI model inputs a prompt sentence to generate an appropriate answer. As a specific example, a "positive prompt sentence" is used to generate an answer with a positive tone. The output is the generated answer.

[0752] Step 11:

[0753] The server quality checks the generated answer to ensure there are no problems with grammar or content. The input is the answer generated in step 10. A quality check module is run to check the answer for accuracy and appropriateness. The output is the final answer that has passed the quality check.

[0754] Step 12:

[0755] The server sends the final answer to the device, which displays the answer to the user. The input is the final answer that passed the quality check in step 11. The server uses the Send API to send the answer to the device and updates the device's UI elements to display it to the user. The output is the answer displayed on the device in a form that the user can see.

[0756] (Application example 2)

[0757] Next, a description will be given of Application 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."

[0758] Conventional generative AI systems have not adequately considered the reliability of content and user sentiment, resulting in a poor user experience and inaccurate information provided. Furthermore, the use of unreliable information can lead to problems such as the provision of inaccurate information and the spread of fake news.

[0759] 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.

[0760] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the content whose reliability has been confirmed in a database, means for training a generative AI model based on the content in the database, means for recognizing and analyzing a user's emotions, means for adjusting the generated answers based on the analyzed emotions, means for quality checking the generated answers and checking for problems with grammar and content, and means for using a public API and secure data transfer means. This makes it possible to provide highly reliable information and optimal information tailored to the user's emotions.

[0761] A "generative AI model" is an algorithm that learns from collected data and generates new information and answers.

[0762] A "certification body" is an organization whose role is to verify and certify the reliability and accuracy of content.

[0763] "Confirmed reliable content" is information whose accuracy and integrity have been confirmed by a certification body.

[0764] A "digital signature" is an electronic means of verifying the origin of content or data and ensuring that it has not been tampered with.

[0765] "Certificate of Accreditation" means official evidence of an accreditation body's confirmation of trustworthiness.

[0766] An "emotion engine" is software that analyzes and classifies emotions from user input text and voice.

[0767] A "database" is a system that organizes and stores collected content and data.

[0768] "Quality check" is the process of checking the generated content for grammar and content appropriateness.

[0769] A "public API" is an interface designed to allow external access to specific functions or data.

[0770] "Secure data transfer methods" are technologies and protocols that prevent unauthorized access or tampering during data transmission.

[0771] A "prompt" is text that instructs a generative AI model on the format and content of input data.

[0772] This invention is a system that combines a generative AI system that uses content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. Specific embodiments of this system are described below.

[0773] Setup and Basic Operation

[0774] The main components of the system are the server, the terminals, and the users, each of which plays a specific role.

[0775] server

[0776] The server is responsible for collecting, validating, storing, learning, generating, and distributing data. Specifically, it works as follows:

[0777] 1. Content collection and verification: The server collects content that has been verified as trustworthy from a trusted authority, using public APIs and secure data transfer methods. The collected content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[0778] 2. Storage and classification in a database: Once the content has been verified as reliable, it is stored in a database. The stored content is analyzed and classified according to its format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[0779] 3. Training the generative AI model: The generative AI model is trained using the shaped dataset. The generative AI model learns patterns and relationships from large amounts of data and has the ability to generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[0780] 4. Request analysis and answer generation: Upon receiving a request from a user, the server analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the generative AI model generates an answer, and a quality check is performed to determine the final answer.

[0781] Terminal

[0782] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[0783] User

[0784] Users access the system through a terminal. They input questions or requests for information into the terminal, and the answers from the server are displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[0785] Specific examples

[0786] Suppose a user puts on smart glasses and inputs, "I'd like to know how to relax after a long day at work." This input is sent to the server via the device. The server analyzes the input, and its emotion engine classifies the user's emotional state as "fatigue." As a result, the server searches for appropriate product information (e.g., aroma diffuser or relaxation massager) from a trusted product database, and generates the following suggestion using a generative AI model: "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or relaxation massager. This aroma diffuser, in particular, has a lavender scent that is expected to promote sound sleep."

[0787] Prompt Sentence Examples

[0788] 1. Prompt to analyze emotions from user input:

[0789] User input text: "I want to know how to relax after a long day at work." Analyze the user's emotional state from this text.

[0790] 2. Prompt to generate sentiment-based product suggestions:

[0791] The user's emotional state has been analyzed as "fatigue." Please generate a sentence to explain to the user about a reliable aroma diffuser or relaxing massager.

[0792] In this way, users receive reliable information in a manner appropriate to their emotional state.

[0793] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0794] Step 1:

[0795] The device receives a user's question or request. The device receives input data by inputting a specific question or request in voice or text format. An example input might be a request such as, "I'd like to know how to relax after a long day at work." The input data is then sent directly to the next step.

[0796] Step 2:

[0797] The device sends the received user question or request to the server, using an Internet connection to transmit the user's input data to the server, which then receives the requested data.

[0798] Step 3:

[0799] The server analyzes the received request. Natural language processing technology is used to extract the intent of the question and key keywords from the text. For example, keywords such as "how to relax" and "long hours at work" may be extracted. Based on the results of this analysis, the system proceeds to the next processing step.

[0800] Step 4:

[0801] The server uses the emotion engine to analyze the emotion of the user's input text. For example, if the user's input is "I want to know how to relax after a long day of work," the emotion engine will classify the user's emotional state as "fatigue" from this text. The analysis result is saved as an emotion category.

[0802] Step 5:

[0803] The server searches for relevant information from a trusted content database based on the analyzed emotional state and key keywords. For example, it searches for content related to "how to relax" or "fatigue" and retrieves the results. These results are then used.

[0804] Step 6:

[0805] The server generates an answer based on the search results using a generative AI model. The generative AI model uses patterns and relationships learned from large amounts of data to generate an appropriate answer. For example, a sentence such as "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or a relaxing massage machine" is generated. This generated sentence is then used.

[0806] Step 7:

[0807] The server then performs a quality check on the generated answer, checking for grammar and content issues and making corrections as necessary. Once the answer passes this check, it is used.

[0808] Step 8:

[0809] The server then adjusts the generated answer based on the results of the emotion analysis. To match the tired emotion, the server may change the tone to something gentler and more relaxing. The final answer is then sent in the next step.

[0810] Step 9:

[0811] The server sends the final answer to the device, using an Internet connection to transmit the answer data to the device, which can then receive the data from the server.

[0812] Step 10:

[0813] The device then displays the received response to the user. The response may be in the form of text or voice, and is delivered in an appropriate tone and expression. For example, a message such as "Thank you for your long hours at work. To help you relax, we recommend using an aroma diffuser or a relaxing massager" may be displayed.

[0814] 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.

[0815] 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> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[0816] 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.

[0817] [Third embodiment]

[0818] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0819] 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.

[0820] 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).

[0821] 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.

[0822] 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.

[0823] 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).

[0824] 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.

[0825] 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.

[0826] 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.

[0827] 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.

[0828] 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.

[0829] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0830] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body. The purpose of this system is to provide reliable information and prevent the spread of fake news and hoaxes. A specific embodiment of this system is described below.

[0831] System Overview

[0832] This system involves a server, a terminal, and a user working together. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[0833] Program processing

[0834] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[0835] server

[0836] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[0837] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[0838] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[0839] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. Based on this, the server searches for relevant content in the database and extracts reliable information.

[0840] Based on the extracted information, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. Once the final answer is determined, the server sends it to the device.

[0841] Terminal

[0842] The terminal acts as the user's interface: the user enters a question or request, which the terminal sends to the server, and upon receiving the answer from the server, the terminal displays the answer to the user.

[0843] User

[0844] Users access the system through terminals, which they type into the terminal questions or requests for information, which then transmit the requests to the server, and the server's responses are displayed on the terminal.

[0845] Specific examples

[0846] Example 1: Medical information question

[0847] User Input

[0848] The user inputs the question "How can I relieve cold symptoms?" into the terminal.

[0849] Server Processing

[0850] The server analyzes the question and searches a database for relevant and reliable medical information. Based on information from certified medical institutions, a generative AI model generates an appropriate answer.

[0851] answer

[0852] The server sends an answer such as, "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays the answer to the user.

[0853] This invention allows users to obtain reliable information and prevents the spread of fake news and rumors.

[0854] The processing flow will be explained below.

[0855] Step 1: Please give me some.

[0856] Example 1

[0857] Next, a description will be given of 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."

[0858] Today, the internet is flooded with unreliable information and fake news, making it difficult for users to quickly obtain reliable information. This problem is particularly severe in specialized fields such as medicine and law, where false information can have a significant impact. Furthermore, conventional systems do not fully establish procedures for verifying reliability, formatting data, and checking the quality of generated answers, meaning the reliability of the information provided to users cannot be guaranteed. To address these issues, a generative AI system is needed that can provide reliable information and prevent the spread of fake news and rumours.

[0859] 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.

[0860] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the verified content in a database, means for analyzing the stored content, classifying it according to format and content, deleting unnecessary data and duplicate data, and formatting it into a dataset, means for training a generative AI model based on the formatted dataset in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed request, means for the generative AI model to generate an answer based on the search results, and means for quality checking the generated answer to check for problems with grammar and content. This allows users to obtain reliable information quickly and accurately.

[0861] A "certification body" is an organization that verifies and certifies the accuracy and reliability of specific content or information with the aim of providing reliable information.

[0862] A "digital signature" is an electronic means of authentication that uses cryptographic technology to verify the identity of the sender of electronic information and whether the information has been tampered with.

[0863] A "certificate of accreditation" is a paper or digital certificate issued by an accreditation body to officially certify the authenticity or accuracy of particular content.

[0864] A "database" is a system for efficiently and systematically storing and managing information and data.

[0865] A "generative AI model" is an artificial intelligence algorithm that learns patterns and relationships from large amounts of data and generates new information and answers.

[0866] A "user request" is a question or request that a user sends to the system for specific information.

[0867] "Natural language processing" is the technology that enables computers to understand, analyze, and generate human language.

[0868] "Quality check" is an inspection process to check whether there are any problems with the grammar or content of the generated answer and to provide it to the user in the most optimal form.

[0869] A "public API" is a public application programming interface that provides an interface for external developers to access specified functionality and data.

[0870] A "secure data transfer method" is a data transfer method that uses encryption and authentication to prevent information leaks and unauthorized access when sending and receiving data.

[0871] This invention is a generative AI system designed to provide reliable information and prevent the spread of fake news and hoaxes. This system operates in cooperation with a server, terminals, and users.

[0872] The server performs processing using the following hardware and software:

[0873] Hardware: High-performance processor, memory, storage, network interface

[0874] Software: Public APIs, secure data transfer methods (e.g., TLS), digital signature verification tools (e.g., OpenSSL), databases (e.g., MySQL), analytics scripts (e.g., Python and the pandas library), generative AI models (e.g., TensorFlow), and natural language processing tools (e.g., Google Cloud Natural Language API and Grammarly API).

[0875] 1. Content Collection

[0876] The server runs a scheduled task to collect trusted content from trusted organizations and sends HTTP GET requests to designated API endpoints. For example, to retrieve medical information, the server retrieves data from the API of a trusted medical institution.

[0877] 2. Reliability check

[0878] The server verifies the digital signature and authentication certificate of the retrieved content. This process uses "OpenSSL" to verify whether the content signature is correct, for example, by using the "openssl verify" command.

[0879] 3. Saving to the database and formatting

[0880] Once the content is verified as reliable, it is stored in a MySQL database. Python scripts are then run to parse the data, classify it by format and content, and clean it up by removing unnecessary and duplicate data, allowing for efficient searching and learning.

[0881] 4. Training the generative AI model

[0882] The resulting dataset is used to train a generative AI model using the TensorFlow library, and each time new data is added to the system, the model is retrained to improve its accuracy.

[0883] 5. User Request Analysis

[0884] The server receives questions sent by users from their devices and analyzes them using natural language processing technology. The server uses the Google Cloud Natural Language API to extract intent and keywords.

[0885] 6. Extraction of reliable information and generation of answers

[0886] The server searches for relevant content in the database and generates an appropriate answer using a generative AI model. This generative AI model generates an answer based on a "prompt sentence." For example, if a user enters "How can I relieve cold symptoms?", the prompt sentence is set to "Provide appropriate medical information based on the user's question."

[0887] 7. Quality check of answers

[0888] The generated answers are quality checked using the Grammarly API to ensure there are no problems with grammar or content.

[0889] 8. Submitting and Viewing Your Answers

[0890] The final answer is sent from the server to the device as an HTTP response, and the device displays it to the user, who can check the answer through the device's interface.

[0891] Specific examples

[0892] Example: Questions about medical information

[0893] User input:

[0894] The user types into the terminal, "How can I relieve my cold symptoms?"

[0895] Server Action:

[0896] The server analyzes the question using the Google Cloud Natural Language API and extracts keywords (e.g., "cold," "symptoms," "relieve").

[0897] The server searches the database for relevant and reliable medical information.

[0898] The generative AI model uses the prompt "How to relieve cold symptoms" to generate an answer.

[0899] The generated answers are quality checked using the Grammarly API.

[0900] answer:

[0901] The server sends the answer "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays it to the user.

[0902] This invention allows users to obtain reliable information quickly and accurately, and eliminates the influence of fake news and hoaxes.

[0903] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0904] Step 1: Content Collection

[0905] The server collects trusted content from a certification authority. As input, it uses the public API endpoint URL. Specifically, the server runs a scheduled task (e.g., a cron job) to send an HTTP GET request. The output is the retrieved content data.

[0906] Step 2: Authenticity check

[0907] The server verifies the digital signature and authentication certificate of the retrieved content. It uses the retrieved content data and its digital signature as input. Specifically, the server uses the "OpenSSL" library to verify the digital signature. The output is the content, which has been confirmed to be authentic.

[0908] Step 3: Saving to the database and formatting

[0909] The server stores the verified content in a database and formats it. The verified content data is used as input. Specifically, the server stores the data in a MySQL database, runs a Python script to analyze the data, and classifies it according to format and content. It then removes unnecessary and duplicate data and formats it into a dataset. The output is a formatted dataset.

[0910] Step 4: Training the generative AI model

[0911] The server trains the generative AI model using the formatted dataset. The formatted dataset is used as input. Specifically, the server retrains the generative AI model using the TensorFlow library. The output is a trained generative AI model.

[0912] Step 5: Receiving a User Request

[0913] The terminal receives requests from the user. As input, it uses questions or requests that the user types into the terminal. Specifically, the terminal receives input through a user interface and sends it to the server. The output is the user request sent to the server.

[0914] Step 6: Parsing the user request

[0915] The server receives the user request and analyzes it using natural language processing technology. The user request is used as input. Specifically, the server uses the Google Cloud Natural Language API to extract the intent and keywords of the request. The output is the analyzed intent and keywords.

[0916] Step 7: Extract reliable information

[0917] The server searches for relevant content in a database based on the parsed intent and keywords. It uses the parsed intent, keywords, and database as input. Specifically, the server executes SQL queries to extract relevant information. The output is the relevant content.

[0918] Step 8: Answer Generation

[0919] The server uses a generative AI model to generate an appropriate answer based on related content. The server uses related content and the generative AI model as input. Specifically, it inputs a "prompt sentence" into the generative AI model to generate an answer. The output is the generated answer. For example, the prompt sentence could be "Provide appropriate medical information based on the user's question."

[0920] Step 9: Check the quality of your answers

[0921] The server checks the quality of the generated answer. It uses the generated answer as input. Specifically, the server checks the grammar and content using the Grammarly API. The output is a quality-checked answer.

[0922] Step 10: Submit and view your responses

[0923] The server sends the quality-checked answer to the terminal, which displays it to the user. The quality-checked answer is used as input. Specifically, the server generates an HTTP response and sends it to the terminal. The terminal analyzes the response data and displays it in its user interface. The output is the answer displayed to the user.

[0924] (Application example 1)

[0925] Next, a description will be given of Application 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."

[0926] In today's information society, the spread of fake news and inaccurate information has become a serious problem. This increases the risk that users will make incorrect decisions based on unreliable information. In particular, unreliable information can easily spread on web pages and social media, so there is a need for a mechanism to quickly detect this and notify users.

[0927] 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.

[0928] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and authentication certificates of the collected content, means for storing the confirmed reliability content in a database, means for training a generative AI model based on the content in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed requests, means for the generative AI model to generate an answer based on the search results, means for evaluating the reliability of content on web pages and social media that users are viewing in real time, and means for displaying warnings for unreliable content. This allows users to make decisions based on reliable information and prevents the spread of fake news and hoaxes.

[0929] An "accreditation body" is an organization or group whose role is to officially verify and certify the reliability of information.

[0930] "Authenticated content" means information or data that has been formally verified and certified for accuracy and authenticity by an accredited organization.

[0931] A "digital signature" is an electronic certificate that is given to ensure the authenticity and integrity of content.

[0932] A "certificate of authenticity" is an official certification that particular content comes from a trusted source.

[0933] A "database" is a collection of information that stores collected content in an organized manner and can be efficiently searched and used.

[0934] A "generative AI model" is an artificial intelligence model that has the ability to learn patterns and relationships from large amounts of data and generate new information and answers.

[0935] A "request" is an inquiry or request that a user sends to the system for information.

[0936] "Analysis" is the process of extracting the intent and keywords of the user's request and understanding their meaning.

[0937] "Related content" is information or data that is relevant to the user's request.

[0938] An "answer" is appropriate information or a solution generated based on a user request.

[0939] A "web page" is a collection of information or data that can be viewed on the Internet.

[0940] "Social media" refers to online platforms that allow users to share information and communicate over the internet.

[0941] "Real-time assessment of trustworthiness" refers to instantly determining the trustworthiness of content as users view it.

[0942] "Displaying a warning" means that when a user encounters unreliable content, a message is displayed to warn the user that there is a problem with the reliability of the information.

[0943] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body, and is capable of evaluating the reliability of web pages and social media content viewed by users in real time and displaying warnings. This system operates in cooperation with a server, terminals, and users.

[0944] System Overview

[0945] The server is primarily responsible for collecting, verifying, storing, learning, generating, and distributing data, while the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[0946] Server Features

[0947] 1. Content Aggregation: The server periodically retrieves trusted content from authorized sources, using public APIs and secure data transfer methods.

[0948] 2. Content verification: The retrieved content is digitally signed and authenticated, and the server verifies this to confirm authenticity.

[0949] 3. Data storage: The validated content is stored in a database, categorized according to format and content, and formatted as a dataset after removing unnecessary and duplicate data.

[0950] 4. Training a generative AI model: The shaped dataset is used to train a generative AI model, which learns patterns and relationships from large amounts of data and generates new information and answers.

[0951] 5. User request analysis: Receive and analyze user requests, using natural language processing technology to extract the intent of the question and keywords.

[0952] 6. Content search: Search for relevant content in the database based on the analysis results and extract reliable information.

[0953] 7. Answer generation: The generative AI model generates an appropriate answer based on the extracted information. The generated answer undergoes a quality check to ensure there are no problems with grammar or content.

[0954] 8. Trustworthiness Assessment: The server assesses the trustworthiness of web pages and social media content viewed by users in real time and generates a warning if the trustworthiness is low.

[0955] Device Features

[0956] The terminal primarily functions as a user interface. When the user enters a question or request, the terminal sends it to the server. When the terminal receives a response from the server, it displays the response to the user. It also displays warnings about unreliable content.

[0957] User operations

[0958] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the server's response is displayed on the terminal. In addition, the reliability of the content the user views is evaluated in real time, and a warning is displayed if the content is unreliable.

[0959] Hardware and software used

[0960] Hardware: Smartphone, Head-Mounted Display (HMD)

[0961] Frontend: React Native (Mobile App Development)

[0962] Backend: Python (Flask)

[0963] AI model: TensorFlow

[0964] Database: PostgreSQL

[0965] NLP tools: nltk, spaCy

[0966] Prompt Sentence Examples

[0967] python

[0968] Example prompt sentence:

[0969] user_input = "View the latest information about COVID-19 vaccines."

[0970] valid_sources = ["CDC", "WHO", "Ministry of Health, Labour and Welfare of Japan"]

[0971] validity_check_prompt = f"Evaluate the authenticity of the information you entered ('{user_input}') based only on data from trusted authorities ({', '.join(valid_sources)})."

[0972] This invention allows users to obtain reliable information and prevents the spread of fake news and rumours. The system's real-time evaluation and warning functions allow users to use information with peace of mind.

[0973] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0974] Step 1:

[0975] Content Collection:

[0976] The server collects content whose authenticity has been confirmed from the certification authority. Specifically, the server periodically obtains certified content using a public API or secure data transfer method. The input is content data from the certification authority, and the output is the collected content stored in the server's local storage or database.

[0977] Step 2:

[0978] Reliability verification:

[0979] The server verifies the digital signature and authentication certificate of the collected content. Specifically, it uses public key cryptography to verify the integrity of the digital signature. As a result, the input is the collected content data, and the output is content whose authenticity has been confirmed.

[0980] Step 3:

[0981] Data Retention and Classification:

[0982] The server stores the verified content in a database, classifies it according to its format and content, removes unnecessary and duplicate data, and formats it into a dataset. The input is verified content, and the output is a formatted dataset.

[0983] Step 4:

[0984] Training generative AI models:

[0985] The server uses the formatted dataset to train a generative AI model, a process that uses frameworks such as TensorFlow to learn patterns and relationships. The input is the formatted dataset, and the output is a trained generative AI model.

[0986] Step 5:

[0987] Receiving and parsing user requests:

[0988] The server receives the user request and analyzes it using natural language processing techniques (such as nltk or spaCy). The input is the user request, and the output is the analyzed question intent and keywords.

[0989] Step 6:

[0990] Search for related content:

[0991] The server searches for relevant content in the database based on the analysis results. Specifically, it uses SQL queries to extract relevant information. The input is the analyzed keywords, and the output is the related content data.

[0992] Step 7:

[0993] Answer generation:

[0994] The server generates an appropriate answer using a generative AI model based on the related content data. The generated answer undergoes quality checks (grammar and content verification) before the final answer is determined. The input is the related content data, and the output is the final answer provided to the user.

[0995] Step 8:

[0996] Reliability Rating and Warnings:

[0997] The server evaluates the trustworthiness of the web page or social media content the user is viewing in real time. If it is judged to be untrustworthy, it displays a warning on the device. The input is the content the user is currently viewing, and the output is a warning message.

[0998] 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.

[0999] This invention combines a generative AI system that uses only content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. This system aims to provide appropriate information taking into account the user's emotions, thereby achieving even higher reliability and user satisfaction. A specific embodiment of this system is described below.

[1000] System Overview

[1001] This system works in cooperation with a server, a terminal, and a user. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal sends user requests to the server and displays the server's responses to the user. The emotion engine also analyzes the emotions in the user's input text and adjusts the content and tone of the response.

[1002] Program processing

[1003] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[1004] server

[1005] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[1006] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[1007] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[1008] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the server searches for relevant content in the database and extracts reliable information.

[1009] Based on the extracted information and the results of emotion analysis, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. The answer is then adjusted based on the user's emotion recognition results, and once the final answer is determined, the server sends it to the device.

[1010] Terminal

[1011] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[1012] User

[1013] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the answer from the server is displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[1014] Specific examples

[1015] Example 1: Nutrition Question

[1016] User Input

[1017] The user inputs the question "What foods are rich in vitamin C?" into the terminal. The user is a little tired and has negative emotions.

[1018] Server Processing

[1019] The server analyzes the question and extracts the keywords "vitamin C" and "food." The emotion engine analyzes the user's emotion and recognizes that it is negative. The server then searches the database for relevant, reliable information, and the generative AI model generates an appropriate answer.

[1020] answer

[1021] The server sends an encouraging response to the device, such as, "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!", and the device displays it to the user.

[1022] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[1023] The processing flow will be explained below.

[1024] Step 1: Content Collection

[1025] server

[1026] Obtain content that is regularly verified as authentic from accredited organizations, using public APIs and secure data transfer methods.

[1027] Step 2: Verify the digital signature and authentication certificate

[1028] server

[1029] Verify the authenticity of the content by verifying the digital signature or authentication certificate attached to the content.

[1030] Step 3: Data storage and analysis

[1031] server

[1032] Content that has been verified as reliable is stored in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[1033] Step 4: Adding Metadata

[1034] server

[1035] Adding metadata to the formatted dataset improves the efficiency of data search and model training.

[1036] Step 5: Training the AI ​​model

[1037] server

[1038] Train a generative AI model using the formatted dataset in the database, then retrain the model as new data is added.

[1039] Step 6: Receiving the request

[1040] Terminal

[1041] The user enters a question or request into the terminal, which then sends the request to the server.

[1042] Step 7: Request Analysis

[1043] server

[1044] Receives requests from users and analyzes them using natural language processing technology to extract the intent of the question and keywords.

[1045] Step 8: Sentiment Analysis

[1046] server

[1047] The emotion engine analyzes the emotion of the user's input text and classifies it into positive, negative, and neutral emotion categories.

[1048] Step 9: Find related content

[1049] server

[1050] Based on the results of request analysis and sentiment analysis, relevant content is searched for in the database.

[1051] Step 10: Generate an answer

[1052] server

[1053] The generative AI model generates appropriate answers to users' questions based on the search results.

[1054] Step 11: Quality check

[1055] server

[1056] Quality check the generated answers to ensure there are no issues with grammar or content.

[1057] Step 12: Emotional Adjustment

[1058] server

[1059] Based on the user's emotion recognition results, the generated responses are adjusted to have an appropriate tone and expression.

[1060] Step 13: Submit your response

[1061] server

[1062] The final answer is sent to the device.

[1063] Step 14: Display to the User

[1064] Terminal

[1065] The answer received from the server is displayed to the user.

[1066] Specific examples

[1067] Example: Nutrition Questions

[1068] User Input

[1069] The user inputs the question "What foods are rich in vitamin C?" into the terminal.

[1070] Step 6: Receiving the request

[1071] The terminal sends a question to the server.

[1072] Step 7: Request Analysis

[1073] The server analyzes the question and extracts the keywords "vitamin C" and "food."

[1074] Step 8: Sentiment Analysis

[1075] The server's emotion engine analyzes the user's input text and recognizes that the user's emotion is negative.

[1076] Step 9: Find related content

[1077] The server retrieves relevant authoritative information from a database.

[1078] Step 10: Generate an answer

[1079] A generative AI model generates the appropriate answer.

[1080] Step 11: Quality check

[1081] The server checks the quality of the generated answers to ensure there are no problems with grammar or content.

[1082] Step 12: Emotional Adjustment

[1083] Add an encouraging tone to your responses to negative emotions.

[1084] Step 13: Submit your response

[1085] The server sends the adjusted response to the terminal.

[1086] Step 14: Display to the User

[1087] The device displays the answer to the user: "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!"

[1088] This process allows users to obtain information that is reliable and delivered in an emotionally appropriate tone, resulting in a more satisfying information experience.

[1089] Example 2

[1090] Next, a description will be given of 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."

[1091] Conventional generative AI systems may provide answers based on unreliable information, putting users at risk of receiving incorrect information. Furthermore, because they do not take user sentiment into consideration, the content and tone of their answers may be inappropriate, resulting in reduced user satisfaction. Furthermore, due to insufficient mechanisms for ensuring data quality and providing reliable information, it is difficult to prevent the spread of fake news and hoaxes.

[1092] 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.

[1093] In this invention, the server includes: means for collecting content whose reliability has been confirmed by a certification authority; means for verifying the digital signature or authentication certificate of the collected content; means for storing the content whose reliability has been confirmed in a storage device; means for training a generative AI model based on the content in the storage device; means for receiving a request from a user and analyzing the request; means for searching for related content in the storage device based on the analyzed request; an emotion engine that analyzes the user's emotions and adjusts the content and tone of a response based on the analysis results; and means for the generative AI model to generate a response based on the search results and the emotion analysis results. This makes it possible to provide highly reliable information in a tone and expression that is appropriate for the user's emotions.

[1094] "Certified Content" means information whose authenticity has been confirmed by a certification body.

[1095] A "digital signature" is a signature that is given electronically and is used to verify the authenticity and authenticity of content.

[1096] "Certificate of Authenticity" means an official certificate issued by a certification authority that indicates that content is trustworthy.

[1097] A "storage device" is a device consisting of hardware and software for storing data.

[1098] A "generative AI model" is an artificial intelligence model that uses machine learning techniques to learn patterns and relationships from large amounts of data and generate new information and answers.

[1099] A "user request" is a question or request for information that a user enters into the system.

[1100] "Natural language processing technology" is a technology for processing human language using a computer, and is used to extract the intent of a question and keywords.

[1101] The "emotion engine" is a function that analyzes the emotions in the user's input text and adjusts the content and tone of the response based on the results.

[1102] "Quality checking" refers to the process of checking generated answers for grammar and content issues.

[1103] "Public API" means an application program interface that is made publicly available for use by external developers and applications.

[1104] A "secure data transfer method" is a data transfer method that employs security measures such as encryption to prevent data eavesdropping or tampering.

[1105] This invention combines a generative AI system that uses content whose authenticity has been confirmed by a certification organization with an emotion engine that recognizes user emotions. The system is designed to operate in cooperation with the server, terminal, and user.

[1106] System configuration

[1107] server

[1108] The server is responsible for collecting, storing, analyzing, generating, and distributing data. Specifically, it operates in the following steps:

[1109] 1. Content collection: The server periodically obtains content whose authenticity has been verified from a certification authority. The server obtains data using public APIs and secure data transfer methods. For example, "https: / / api.certifiedcontent.org / data" is used as the public API.

[1110] 2. Authenticity verification: Verify the authenticity of the retrieved content by verifying its digital signature or authentication certificate.

[1111] 3. Data storage and formatting: Once content is verified as reliable, it is stored on a storage device, categorized by format and content, and redundant and duplicate data is removed. The organized dataset is then given metadata to enable efficient searching and learning.

[1112] 4. Training a generative AI model: A generative AI model is trained using the shaped dataset, using a framework such as TensorFlow or PyTorch, and is retrained every time new data is added.

[1113] Terminal

[1114] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays it to the user. The response is adjusted by the emotion engine and is delivered in a tone and expression that matches the user's emotions.

[1115] User

[1116] Users access the system through a terminal. They input questions or requests for information into the terminal, and the terminal sends the request to the server. The server's response is displayed on the terminal. The user can receive responses tailored by the emotion engine, resulting in a more satisfying experience.

[1117] Specific examples

[1118] Example 1: Nutrition Question

[1119] User Input

[1120] The user inputs the question "What foods are rich in vitamin C?" into the terminal. It is assumed that the user is slightly tired and in a negative emotional state.

[1121] Server Processing

[1122] The server receives the question and uses natural language processing technology to extract keywords such as "vitamin C" and "food." The emotion engine analyzes the user's emotions and recognizes negative emotions. The server then searches the database for relevant and reliable information, and the generative AI model generates an appropriate answer.

[1123] answer

[1124] The server generates a response such as the following and sends it to the device: "Foods that are rich in vitamin C include oranges, kiwis, and bell peppers. Eating these foods will help you stay healthy!" The response has an encouraging tone and is displayed on the device to the user.

[1125] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[1126] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1127] Step 1:

[1128] The server collects content whose authenticity has been confirmed from a certification authority. The input is data obtained through a public API or secure data transfer method. Specifically, the server uses the public API to call "https: / / api.certifiedcontent.org / data" and receives data in JSON format. The output is the raw data before its authenticity is confirmed.

[1129] Step 2:

[1130] The server verifies the digital signatures and authentication certificates of the collected content. The input is the content collected in step 1. Specifically, it verifies its authenticity using an algorithm that verifies the digital signature. The output is the data whose authenticity has been confirmed.

[1131] Step 3:

[1132] The server saves the content whose authenticity has been confirmed in the storage device. The input is the data whose authenticity has been confirmed in step 2. Specifically, it executes the "save" command to store the data in the database. The output is the trusted information as the data stored in the database.

[1133] Step 4:

[1134] The server analyzes the stored content and classifies it according to format and content. The input is the data in the database stored in step 3. It uses queries to classify the data into categories such as "nutrition," "health," and "emotions," and then refines the dataset by removing unnecessary and duplicate data. The output is the refined dataset.

[1135] Step 5:

[1136] The server trains the generative AI model using the formatted dataset. The input is the formatted dataset from step 4. Specifically, it runs a Python script and trains the model using TensorFlow or PyTorch. If necessary, it also retrains the model when new data is added. The output is a trained generative AI model.

[1137] Step 6:

[1138] The user inputs a question or request through the terminal. The input is text data representing the user's question or request. The terminal sends this request to the server. Specific operations include text input and transmission functions through the terminal's UI. The output is the request data sent to the server.

[1139] Step 7:

[1140] The server receives a request from the user and analyzes it. The input is the request data sent in step 6. Specifically, it uses a natural language processing library such as "NLTK" or "spaCy" to extract the intent and keywords of the question. The output is the analyzed intent and keyword information of the question.

[1141] Step 8:

[1142] The server runs an emotion engine that analyzes the emotion of the user's input text. The input is the question intent and keyword information extracted in step 7. The emotion analysis module is used to classify the emotion into positive, negative, and neutral categories. The output is data indicating the user's emotional state.

[1143] Step 9:

[1144] The server searches for relevant content in the database based on the analyzed request and emotional information. The input is the emotional state and question intent / keyword information obtained in step 8. It executes an SQL query to retrieve relevant information. The output is reliable relevant information.

[1145] Step 10:

[1146] The server uses the generative AI model to generate an answer based on the search results and sentiment analysis results. The input is the relevant information and emotional state obtained in step 9. The generative AI model inputs a prompt sentence to generate an appropriate answer. As a specific example, a "positive prompt sentence" is used to generate an answer with a positive tone. The output is the generated answer.

[1147] Step 11:

[1148] The server quality checks the generated answer to ensure there are no problems with grammar or content. The input is the answer generated in step 10. A quality check module is run to check the answer for accuracy and appropriateness. The output is the final answer that has passed the quality check.

[1149] Step 12:

[1150] The server sends the final answer to the device, which displays the answer to the user. The input is the final answer that passed the quality check in step 11. The server uses the Send API to send the answer to the device and updates the device's UI elements to display it to the user. The output is the answer displayed on the device in a form that the user can see.

[1151] (Application example 2)

[1152] Next, a description will be given of Application 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."

[1153] Conventional generative AI systems have not adequately considered the reliability of content and user sentiment, resulting in a poor user experience and inaccurate information provided. Furthermore, the use of unreliable information can lead to problems such as the provision of inaccurate information and the spread of fake news.

[1154] 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.

[1155] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the content whose reliability has been confirmed in a database, means for training a generative AI model based on the content in the database, means for recognizing and analyzing a user's emotions, means for adjusting the generated answers based on the analyzed emotions, means for quality checking the generated answers and checking for problems with grammar and content, and means for using a public API and secure data transfer means. This makes it possible to provide highly reliable information and optimal information tailored to the user's emotions.

[1156] A "generative AI model" is an algorithm that learns from collected data and generates new information and answers.

[1157] A "certification body" is an organization whose role is to verify and certify the reliability and accuracy of content.

[1158] "Confirmed reliable content" is information whose accuracy and integrity have been confirmed by a certification body.

[1159] A "digital signature" is an electronic means of verifying the origin of content or data and ensuring that it has not been tampered with.

[1160] "Certificate of Accreditation" means official evidence of an accreditation body's confirmation of trustworthiness.

[1161] An "emotion engine" is software that analyzes and classifies emotions from user input text and voice.

[1162] A "database" is a system that organizes and stores collected content and data.

[1163] "Quality check" is the process of checking the generated content for grammar and content appropriateness.

[1164] A "public API" is an interface designed to allow external access to specific functions or data.

[1165] "Secure data transfer methods" are technologies and protocols that prevent unauthorized access or tampering during data transmission.

[1166] A "prompt" is text that instructs a generative AI model on the format and content of input data.

[1167] This invention is a system that combines a generative AI system that uses content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. Specific embodiments of this system are described below.

[1168] Setup and Basic Operation

[1169] The main components of the system are the server, the terminals, and the users, each of which plays a specific role.

[1170] server

[1171] The server is responsible for collecting, validating, storing, learning, generating, and distributing data. Specifically, it works as follows:

[1172] 1. Content collection and verification: The server collects content that has been verified as trustworthy from a trusted authority, using public APIs and secure data transfer methods. The collected content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[1173] 2. Storage and classification in a database: Once the content has been verified as reliable, it is stored in a database. The stored content is analyzed and classified according to its format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[1174] 3. Training the generative AI model: The generative AI model is trained using the shaped dataset. The generative AI model learns patterns and relationships from large amounts of data and has the ability to generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[1175] 4. Request analysis and answer generation: Upon receiving a request from a user, the server analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the generative AI model generates an answer, and a quality check is performed to determine the final answer.

[1176] Terminal

[1177] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[1178] User

[1179] Users access the system through a terminal. They input questions or requests for information into the terminal, and the answers from the server are displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[1180] Specific examples

[1181] Suppose a user puts on smart glasses and inputs, "I'd like to know how to relax after a long day at work." This input is sent to the server via the device. The server analyzes the input, and its emotion engine classifies the user's emotional state as "fatigue." As a result, the server searches for appropriate product information (e.g., aroma diffuser or relaxation massager) from a trusted product database, and generates the following suggestion using a generative AI model: "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or relaxation massager. This aroma diffuser, in particular, has a lavender scent that is expected to promote sound sleep."

[1182] Prompt Sentence Examples

[1183] 1. Prompt to analyze emotions from user input:

[1184] User input text: "I want to know how to relax after a long day at work." Analyze the user's emotional state from this text.

[1185] 2. Prompt to generate sentiment-based product suggestions:

[1186] The user's emotional state has been analyzed as "fatigue." Please generate a sentence to explain to the user about a reliable aroma diffuser or relaxing massager.

[1187] In this way, users receive reliable information in a manner appropriate to their emotional state.

[1188] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1189] Step 1:

[1190] The device receives a user's question or request. The device receives input data by inputting a specific question or request in voice or text format. An example input might be a request such as, "I'd like to know how to relax after a long day at work." The input data is then sent directly to the next step.

[1191] Step 2:

[1192] The device sends the received user question or request to the server, using an Internet connection to transmit the user's input data to the server, which then receives the requested data.

[1193] Step 3:

[1194] The server analyzes the received request. Natural language processing technology is used to extract the intent of the question and key keywords from the text. For example, keywords such as "how to relax" and "long hours at work" may be extracted. Based on the results of this analysis, the system proceeds to the next processing step.

[1195] Step 4:

[1196] The server uses the emotion engine to analyze the emotion of the user's input text. For example, if the user's input is "I want to know how to relax after a long day of work," the emotion engine will classify the user's emotional state as "fatigue" from this text. The analysis result is saved as an emotion category.

[1197] Step 5:

[1198] The server searches for relevant information from a trusted content database based on the analyzed emotional state and key keywords. For example, it searches for content related to "how to relax" or "fatigue" and retrieves the results. These results are then used.

[1199] Step 6:

[1200] The server generates an answer based on the search results using a generative AI model. The generative AI model uses patterns and relationships learned from large amounts of data to generate an appropriate answer. For example, a sentence such as "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or a relaxing massage machine" is generated. This generated sentence is then used.

[1201] Step 7:

[1202] The server then performs a quality check on the generated answer, checking for grammar and content issues and making corrections as necessary. Once the answer passes this check, it is used.

[1203] Step 8:

[1204] The server then adjusts the generated answer based on the results of the emotion analysis. To match the tired emotion, the server may change the tone to something gentler and more relaxing. The final answer is then sent in the next step.

[1205] Step 9:

[1206] The server sends the final answer to the device, using an Internet connection to transmit the answer data to the device, which can then receive the data from the server.

[1207] Step 10:

[1208] The device then displays the received response to the user. The response may be in the form of text or voice, and is delivered in an appropriate tone and expression. For example, a message such as "Thank you for your long hours at work. To help you relax, we recommend using an aroma diffuser or a relaxing massager" may be displayed.

[1209] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 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.

[1210] 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> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[1211] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1212] [Fourth embodiment]

[1213] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1214] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1215] 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).

[1216] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. 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 control target 443 are also connected to the bus 52.

[1217] 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.

[1218] 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).

[1219] 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.

[1220] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1221] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, 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.

[1222] 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.

[1223] 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.

[1224] In the robot 414, the processor 46 performs the reception output process. 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.

[1225] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1226] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body. The purpose of this system is to provide reliable information and prevent the spread of fake news and hoaxes. A specific embodiment of this system is described below.

[1227] System Overview

[1228] This system involves a server, a terminal, and a user working together. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[1229] Program processing

[1230] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[1231] server

[1232] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[1233] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[1234] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[1235] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. Based on this, the server searches for relevant content in the database and extracts reliable information.

[1236] Based on the extracted information, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. Once the final answer is determined, the server sends it to the device.

[1237] Terminal

[1238] The terminal acts as the user's interface: the user enters a question or request, which the terminal sends to the server, and upon receiving the answer from the server, the terminal displays the answer to the user.

[1239] User

[1240] Users access the system through terminals, which they type into the terminal questions or requests for information, which then transmit the requests to the server, and the server's responses are displayed on the terminal.

[1241] Specific examples

[1242] Example 1: Medical information question

[1243] User Input

[1244] The user inputs the question "How can I relieve cold symptoms?" into the terminal.

[1245] Server Processing

[1246] The server analyzes the question and searches a database for relevant and reliable medical information. Based on information from certified medical institutions, a generative AI model generates an appropriate answer.

[1247] answer

[1248] The server sends an answer such as, "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays the answer to the user.

[1249] This invention allows users to obtain reliable information and prevents the spread of fake news and rumors.

[1250] The processing flow will be explained below.

[1251] Step 1: Please give me some.

[1252] Example 1

[1253] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1254] Today, the internet is flooded with unreliable information and fake news, making it difficult for users to quickly obtain reliable information. This problem is particularly severe in specialized fields such as medicine and law, where false information can have a significant impact. Furthermore, conventional systems do not fully establish procedures for verifying reliability, formatting data, and checking the quality of generated answers, meaning the reliability of the information provided to users cannot be guaranteed. To address these issues, a generative AI system is needed that can provide reliable information and prevent the spread of fake news and rumours.

[1255] 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.

[1256] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the verified content in a database, means for analyzing the stored content, classifying it according to format and content, deleting unnecessary data and duplicate data, and formatting it into a dataset, means for training a generative AI model based on the formatted dataset in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed request, means for the generative AI model to generate an answer based on the search results, and means for quality checking the generated answer to check for problems with grammar and content. This allows users to obtain reliable information quickly and accurately.

[1257] A "certification body" is an organization that verifies and certifies the accuracy and reliability of specific content or information with the aim of providing reliable information.

[1258] A "digital signature" is an electronic means of authentication that uses cryptographic technology to verify the identity of the sender of electronic information and whether the information has been tampered with.

[1259] A "certificate of accreditation" is a paper or digital certificate issued by an accreditation body to officially certify the authenticity or accuracy of particular content.

[1260] A "database" is a system for efficiently and systematically storing and managing information and data.

[1261] A "generative AI model" is an artificial intelligence algorithm that learns patterns and relationships from large amounts of data and generates new information and answers.

[1262] A "user request" is a question or request that a user sends to the system for specific information.

[1263] "Natural language processing" is the technology that enables computers to understand, analyze, and generate human language.

[1264] "Quality check" is an inspection process to check whether there are any problems with the grammar or content of the generated answer and to provide it to the user in the most optimal form.

[1265] A "public API" is a public application programming interface that provides an interface for external developers to access specified functionality and data.

[1266] A "secure data transfer method" is a data transfer method that uses encryption and authentication to prevent information leaks and unauthorized access when sending and receiving data.

[1267] This invention is a generative AI system designed to provide reliable information and prevent the spread of fake news and hoaxes. This system operates in cooperation with a server, terminals, and users.

[1268] The server performs processing using the following hardware and software:

[1269] Hardware: High-performance processor, memory, storage, network interface

[1270] Software: Public APIs, secure data transfer methods (e.g., TLS), digital signature verification tools (e.g., OpenSSL), databases (e.g., MySQL), analytics scripts (e.g., Python and the pandas library), generative AI models (e.g., TensorFlow), and natural language processing tools (e.g., Google Cloud Natural Language API and Grammarly API).

[1271] 1. Content Collection

[1272] The server runs a scheduled task to collect trusted content from trusted organizations and sends HTTP GET requests to designated API endpoints. For example, to retrieve medical information, the server retrieves data from the API of a trusted medical institution.

[1273] 2. Reliability check

[1274] The server verifies the digital signature and authentication certificate of the retrieved content. This process uses "OpenSSL" to verify whether the content signature is correct, for example, by using the "openssl verify" command.

[1275] 3. Saving to the database and formatting

[1276] Once the content is verified as reliable, it is stored in a MySQL database. Python scripts are then run to parse the data, classify it by format and content, and clean it up by removing unnecessary and duplicate data, allowing for efficient searching and learning.

[1277] 4. Training the generative AI model

[1278] The resulting dataset is used to train a generative AI model using the TensorFlow library, and each time new data is added to the system, the model is retrained to improve its accuracy.

[1279] 5. User Request Analysis

[1280] The server receives questions sent by users from their devices and analyzes them using natural language processing technology. The server uses the Google Cloud Natural Language API to extract intent and keywords.

[1281] 6. Extraction of reliable information and generation of answers

[1282] The server searches for relevant content in the database and generates an appropriate answer using a generative AI model. This generative AI model generates an answer based on a "prompt sentence." For example, if a user enters "How can I relieve cold symptoms?", the prompt sentence is set to "Provide appropriate medical information based on the user's question."

[1283] 7. Quality check of answers

[1284] The generated answers are quality checked using the Grammarly API to ensure there are no problems with grammar or content.

[1285] 8. Submitting and Viewing Your Answers

[1286] The final answer is sent from the server to the device as an HTTP response, and the device displays it to the user, who can check the answer through the device's interface.

[1287] Specific examples

[1288] Example: Questions about medical information

[1289] User input:

[1290] The user types into the terminal, "How can I relieve my cold symptoms?"

[1291] Server Action:

[1292] The server analyzes the question using the Google Cloud Natural Language API and extracts keywords (e.g., "cold," "symptoms," "relieve").

[1293] The server searches the database for relevant and reliable medical information.

[1294] The generative AI model uses the prompt "How to relieve cold symptoms" to generate an answer.

[1295] The generated answers are quality checked using the Grammarly API.

[1296] answer:

[1297] The server sends the answer "To alleviate cold symptoms, it is important to get enough rest and stay hydrated. It is also effective to eat foods rich in vitamin C and maintain appropriate room temperature and humidity," to the device, which then displays it to the user.

[1298] This invention allows users to obtain reliable information quickly and accurately, and eliminates the influence of fake news and hoaxes.

[1299] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1300] Step 1: Content Collection

[1301] The server collects trusted content from a certification authority. As input, it uses the public API endpoint URL. Specifically, the server runs a scheduled task (e.g., a cron job) to send an HTTP GET request. The output is the retrieved content data.

[1302] Step 2: Authenticity check

[1303] The server verifies the digital signature and authentication certificate of the retrieved content. It uses the retrieved content data and its digital signature as input. Specifically, the server uses the "OpenSSL" library to verify the digital signature. The output is the content, which has been confirmed to be authentic.

[1304] Step 3: Saving to the database and formatting

[1305] The server stores the verified content in a database and formats it. The verified content data is used as input. Specifically, the server stores the data in a MySQL database, runs a Python script to analyze the data, and classifies it according to format and content. It then removes unnecessary and duplicate data and formats it into a dataset. The output is a formatted dataset.

[1306] Step 4: Training the generative AI model

[1307] The server trains the generative AI model using the formatted dataset. The formatted dataset is used as input. Specifically, the server retrains the generative AI model using the TensorFlow library. The output is a trained generative AI model.

[1308] Step 5: Receiving a User Request

[1309] The terminal receives requests from the user. As input, it uses questions or requests that the user types into the terminal. Specifically, the terminal receives input through a user interface and sends it to the server. The output is the user request sent to the server.

[1310] Step 6: Parsing the user request

[1311] The server receives the user request and analyzes it using natural language processing technology. The user request is used as input. Specifically, the server uses the Google Cloud Natural Language API to extract the intent and keywords of the request. The output is the analyzed intent and keywords.

[1312] Step 7: Extract reliable information

[1313] The server searches for relevant content in a database based on the parsed intent and keywords. It uses the parsed intent, keywords, and database as input. Specifically, the server executes SQL queries to extract relevant information. The output is the relevant content.

[1314] Step 8: Answer Generation

[1315] The server uses a generative AI model to generate an appropriate answer based on related content. The server uses related content and the generative AI model as input. Specifically, it inputs a "prompt sentence" into the generative AI model to generate an answer. The output is the generated answer. For example, the prompt sentence could be "Provide appropriate medical information based on the user's question."

[1316] Step 9: Check the quality of your answers

[1317] The server checks the quality of the generated answer. It uses the generated answer as input. Specifically, the server checks the grammar and content using the Grammarly API. The output is a quality-checked answer.

[1318] Step 10: Submit and view your responses

[1319] The server sends the quality-checked answer to the terminal, which displays it to the user. The quality-checked answer is used as input. Specifically, the server generates an HTTP response and sends it to the terminal. The terminal analyzes the response data and displays it in its user interface. The output is the answer displayed to the user.

[1320] (Application example 1)

[1321] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1322] In today's information society, the spread of fake news and inaccurate information has become a serious problem. This increases the risk that users will make incorrect decisions based on unreliable information. In particular, unreliable information can easily spread on web pages and social media, so there is a need for a mechanism to quickly detect this and notify users.

[1323] 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.

[1324] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and authentication certificates of the collected content, means for storing the confirmed reliability content in a database, means for training a generative AI model based on the content in the database, means for receiving requests from users and analyzing the requests, means for searching for related content in the database based on the analyzed requests, means for the generative AI model to generate an answer based on the search results, means for evaluating the reliability of content on web pages and social media that users are viewing in real time, and means for displaying warnings for unreliable content. This allows users to make decisions based on reliable information and prevents the spread of fake news and hoaxes.

[1325] An "accreditation body" is an organization or group whose role is to officially verify and certify the reliability of information.

[1326] "Authenticated content" means information or data that has been formally verified and certified for accuracy and authenticity by an accredited organization.

[1327] A "digital signature" is an electronic certificate that is given to ensure the authenticity and integrity of content.

[1328] A "certificate of authenticity" is an official certification that particular content comes from a trusted source.

[1329] A "database" is a collection of information that stores collected content in an organized manner and can be efficiently searched and used.

[1330] A "generative AI model" is an artificial intelligence model that has the ability to learn patterns and relationships from large amounts of data and generate new information and answers.

[1331] A "request" is an inquiry or request that a user sends to the system for information.

[1332] "Analysis" is the process of extracting the intent and keywords of the user's request and understanding their meaning.

[1333] "Related content" is information or data that is relevant to the user's request.

[1334] An "answer" is appropriate information or a solution generated based on a user request.

[1335] A "web page" is a collection of information or data that can be viewed on the Internet.

[1336] "Social media" refers to online platforms that allow users to share information and communicate over the internet.

[1337] "Real-time assessment of trustworthiness" refers to instantly determining the trustworthiness of content as users view it.

[1338] "Displaying a warning" means that when a user encounters unreliable content, a message is displayed to warn the user that there is a problem with the reliability of the information.

[1339] This invention is a generative AI system that uses only content whose reliability has been confirmed by a certification body, and is capable of evaluating the reliability of web pages and social media content viewed by users in real time and displaying warnings. This system operates in cooperation with a server, terminals, and users.

[1340] System Overview

[1341] The server is primarily responsible for collecting, verifying, storing, learning, generating, and distributing data, while the terminal is responsible for sending user requests to the server and displaying the server's responses to the user.

[1342] Server Features

[1343] 1. Content Aggregation: The server periodically retrieves trusted content from authorized sources, using public APIs and secure data transfer methods.

[1344] 2. Content verification: The retrieved content is digitally signed and authenticated, and the server verifies this to confirm authenticity.

[1345] 3. Data storage: The validated content is stored in a database, categorized according to format and content, and formatted as a dataset after removing unnecessary and duplicate data.

[1346] 4. Training a generative AI model: The shaped dataset is used to train a generative AI model, which learns patterns and relationships from large amounts of data and generates new information and answers.

[1347] 5. User request analysis: Receive and analyze user requests, using natural language processing technology to extract the intent of the question and keywords.

[1348] 6. Content search: Search for relevant content in the database based on the analysis results and extract reliable information.

[1349] 7. Answer generation: The generative AI model generates an appropriate answer based on the extracted information. The generated answer undergoes a quality check to ensure there are no problems with grammar or content.

[1350] 8. Trustworthiness Assessment: The server assesses the trustworthiness of web pages and social media content viewed by users in real time and generates a warning if the trustworthiness is low.

[1351] Device Features

[1352] The terminal primarily functions as a user interface. When the user enters a question or request, the terminal sends it to the server. When the terminal receives a response from the server, it displays the response to the user. It also displays warnings about unreliable content.

[1353] User operations

[1354] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the server's response is displayed on the terminal. In addition, the reliability of the content the user views is evaluated in real time, and a warning is displayed if the content is unreliable.

[1355] Hardware and software used

[1356] Hardware: Smartphone, Head-Mounted Display (HMD)

[1357] Frontend: React Native (Mobile App Development)

[1358] Backend: Python (Flask)

[1359] AI model: TensorFlow

[1360] Database: PostgreSQL

[1361] NLP tools: nltk, spaCy

[1362] Prompt Sentence Examples

[1363] python

[1364] Example prompt sentence:

[1365] user_input = "View the latest information about COVID-19 vaccines."

[1366] valid_sources = ["CDC", "WHO", "Ministry of Health, Labour and Welfare of Japan"]

[1367] validity_check_prompt = f"Evaluate the authenticity of the information you entered ('{user_input}') based only on data from trusted authorities ({', '.join(valid_sources)})."

[1368] This invention allows users to obtain reliable information and prevents the spread of fake news and rumours. The system's real-time evaluation and warning functions allow users to use information with peace of mind.

[1369] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1370] Step 1:

[1371] Content Collection:

[1372] The server collects content whose authenticity has been confirmed from the certification authority. Specifically, the server periodically obtains certified content using a public API or secure data transfer method. The input is content data from the certification authority, and the output is the collected content stored in the server's local storage or database.

[1373] Step 2:

[1374] Reliability verification:

[1375] The server verifies the digital signature and authentication certificate of the collected content. Specifically, it uses public key cryptography to verify the integrity of the digital signature. As a result, the input is the collected content data, and the output is content whose authenticity has been confirmed.

[1376] Step 3:

[1377] Data Retention and Classification:

[1378] The server stores the verified content in a database, classifies it according to its format and content, removes unnecessary and duplicate data, and formats it into a dataset. The input is verified content, and the output is a formatted dataset.

[1379] Step 4:

[1380] Training generative AI models:

[1381] The server uses the formatted dataset to train a generative AI model, a process that uses frameworks such as TensorFlow to learn patterns and relationships. The input is the formatted dataset, and the output is a trained generative AI model.

[1382] Step 5:

[1383] Receiving and parsing user requests:

[1384] The server receives the user request and analyzes it using natural language processing techniques (such as nltk or spaCy). The input is the user request, and the output is the analyzed question intent and keywords.

[1385] Step 6:

[1386] Search for related content:

[1387] The server searches for relevant content in the database based on the analysis results. Specifically, it uses SQL queries to extract relevant information. The input is the analyzed keywords, and the output is the related content data.

[1388] Step 7:

[1389] Answer generation:

[1390] The server generates an appropriate answer using a generative AI model based on the related content data. The generated answer undergoes quality checks (grammar and content verification) before the final answer is determined. The input is the related content data, and the output is the final answer provided to the user.

[1391] Step 8:

[1392] Reliability Rating and Warnings:

[1393] The server evaluates the trustworthiness of the web page or social media content the user is viewing in real time. If it is judged to be untrustworthy, it displays a warning on the device. The input is the content the user is currently viewing, and the output is a warning message.

[1394] 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.

[1395] This invention combines a generative AI system that uses only content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. This system aims to provide appropriate information taking into account the user's emotions, thereby achieving even higher reliability and user satisfaction. A specific embodiment of this system is described below.

[1396] System Overview

[1397] This system works in cooperation with a server, a terminal, and a user. The server is responsible for collecting, verifying, storing, learning, generating, and distributing data. Meanwhile, the terminal sends user requests to the server and displays the server's responses to the user. The emotion engine also analyzes the emotions in the user's input text and adjusts the content and tone of the response.

[1398] Program processing

[1399] The processes that each server, terminal, and user is responsible for will be explained in natural language below.

[1400] server

[1401] The server acts as a content aggregator, periodically obtaining trusted content from authorized sources using public APIs and secure data transfer methods. The content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[1402] Once the authenticity has been confirmed, the server stores the content in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is also removed, and the data is organized into a dataset. This allows metadata to be added to the dataset, improving search and learning efficiency.

[1403] The server then uses the shaped dataset to train a generative AI model, which can learn patterns and relationships from large amounts of data and generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[1404] When a user sends a specific question or request from their device, the server receives and analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the server searches for relevant content in the database and extracts reliable information.

[1405] Based on the extracted information and the results of emotion analysis, the generative AI model generates an appropriate answer to the user's question. The generated answer undergoes a quality check to ensure there are no problems with grammar or content. The answer is then adjusted based on the user's emotion recognition results, and once the final answer is determined, the server sends it to the device.

[1406] Terminal

[1407] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[1408] User

[1409] Users access the system through a terminal. They input questions or requests for information into the terminal. The terminal sends the request to the server, and the answer from the server is displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[1410] Specific examples

[1411] Example 1: Nutrition Question

[1412] User Input

[1413] The user inputs the question "What foods are rich in vitamin C?" into the terminal. The user is a little tired and has negative emotions.

[1414] Server Processing

[1415] The server analyzes the question and extracts the keywords "vitamin C" and "food." The emotion engine analyzes the user's emotion and recognizes that it is negative. The server then searches the database for relevant, reliable information, and the generative AI model generates an appropriate answer.

[1416] answer

[1417] The server sends an encouraging response to the device, such as, "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!", and the device displays it to the user.

[1418] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[1419] The processing flow will be explained below.

[1420] Step 1: Content Collection

[1421] server

[1422] Obtain content that is regularly verified as authentic from accredited organizations, using public APIs and secure data transfer methods.

[1423] Step 2: Verify the digital signature and authentication certificate

[1424] server

[1425] Verify the authenticity of the content by verifying the digital signature or authentication certificate attached to the content.

[1426] Step 3: Data storage and analysis

[1427] server

[1428] Content that has been verified as reliable is stored in a database. The stored content is analyzed and classified according to format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[1429] Step 4: Adding Metadata

[1430] server

[1431] Adding metadata to the formatted dataset improves the efficiency of data search and model training.

[1432] Step 5: Training the AI ​​model

[1433] server

[1434] Train a generative AI model using the formatted dataset in the database, then retrain the model as new data is added.

[1435] Step 6: Receiving the request

[1436] Terminal

[1437] The user enters a question or request into the terminal, which then sends the request to the server.

[1438] Step 7: Request Analysis

[1439] server

[1440] Receives requests from users and analyzes them using natural language processing technology to extract the intent of the question and keywords.

[1441] Step 8: Sentiment Analysis

[1442] server

[1443] The emotion engine analyzes the emotion of the user's input text and classifies it into positive, negative, and neutral emotion categories.

[1444] Step 9: Find related content

[1445] server

[1446] Based on the results of request analysis and sentiment analysis, relevant content is searched for in the database.

[1447] Step 10: Generate an answer

[1448] server

[1449] The generative AI model generates appropriate answers to users' questions based on the search results.

[1450] Step 11: Quality check

[1451] server

[1452] Quality check the generated answers to ensure there are no issues with grammar or content.

[1453] Step 12: Emotional Adjustment

[1454] server

[1455] Based on the user's emotion recognition results, the generated responses are adjusted to have an appropriate tone and expression.

[1456] Step 13: Submit your response

[1457] server

[1458] The final answer is sent to the device.

[1459] Step 14: Display to the User

[1460] Terminal

[1461] The answer received from the server is displayed to the user.

[1462] Specific examples

[1463] Example: Nutrition Questions

[1464] User Input

[1465] The user inputs the question "What foods are rich in vitamin C?" into the terminal.

[1466] Step 6: Receiving the request

[1467] The terminal sends a question to the server.

[1468] Step 7: Request Analysis

[1469] The server analyzes the question and extracts the keywords "vitamin C" and "food."

[1470] Step 8: Sentiment Analysis

[1471] The server's emotion engine analyzes the user's input text and recognizes that the user's emotion is negative.

[1472] Step 9: Find related content

[1473] The server retrieves relevant authoritative information from a database.

[1474] Step 10: Generate an answer

[1475] A generative AI model generates the appropriate answer.

[1476] Step 11: Quality check

[1477] The server checks the quality of the generated answers to ensure there are no problems with grammar or content.

[1478] Step 12: Emotional Adjustment

[1479] Add an encouraging tone to your responses to negative emotions.

[1480] Step 13: Submit your response

[1481] The server sends the adjusted response to the terminal.

[1482] Step 14: Display to the User

[1483] The device displays the answer to the user: "Foods that are high in vitamin C include oranges, kiwis, and bell peppers. Eating these foods can help you maintain your health!"

[1484] This process allows users to obtain information that is reliable and delivered in an emotionally appropriate tone, resulting in a more satisfying information experience.

[1485] Example 2

[1486] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1487] Conventional generative AI systems may provide answers based on unreliable information, putting users at risk of receiving incorrect information. Furthermore, because they do not take user sentiment into consideration, the content and tone of their answers may be inappropriate, resulting in reduced user satisfaction. Furthermore, due to insufficient mechanisms for ensuring data quality and providing reliable information, it is difficult to prevent the spread of fake news and hoaxes.

[1488] 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.

[1489] In this invention, the server includes: means for collecting content whose reliability has been confirmed by a certification authority; means for verifying the digital signature or authentication certificate of the collected content; means for storing the content whose reliability has been confirmed in a storage device; means for training a generative AI model based on the content in the storage device; means for receiving a request from a user and analyzing the request; means for searching for related content in the storage device based on the analyzed request; an emotion engine that analyzes the user's emotions and adjusts the content and tone of a response based on the analysis results; and means for the generative AI model to generate a response based on the search results and the emotion analysis results. This makes it possible to provide highly reliable information in a tone and expression that is appropriate for the user's emotions.

[1490] "Certified Content" means information whose authenticity has been confirmed by a certification body.

[1491] A "digital signature" is a signature that is given electronically and is used to verify the authenticity and authenticity of content.

[1492] "Certificate of Authenticity" means an official certificate issued by a certification authority that indicates that content is trustworthy.

[1493] A "storage device" is a device consisting of hardware and software for storing data.

[1494] A "generative AI model" is an artificial intelligence model that uses machine learning techniques to learn patterns and relationships from large amounts of data and generate new information and answers.

[1495] A "user request" is a question or request for information that a user enters into the system.

[1496] "Natural language processing technology" is a technology for processing human language using a computer, and is used to extract the intent of a question and keywords.

[1497] The "emotion engine" is a function that analyzes the emotions in the user's input text and adjusts the content and tone of the response based on the results.

[1498] "Quality checking" refers to the process of checking generated answers for grammar and content issues.

[1499] "Public API" means an application program interface that is made publicly available for use by external developers and applications.

[1500] A "secure data transfer method" is a data transfer method that employs security measures such as encryption to prevent data eavesdropping or tampering.

[1501] This invention combines a generative AI system that uses content whose authenticity has been confirmed by a certification organization with an emotion engine that recognizes user emotions. The system is designed to operate in cooperation with the server, terminal, and user.

[1502] System configuration

[1503] server

[1504] The server is responsible for collecting, storing, analyzing, generating, and distributing data. Specifically, it operates in the following steps:

[1505] 1. Content collection: The server periodically obtains content whose authenticity has been verified from a certification authority. The server obtains data using public APIs and secure data transfer methods. For example, "https: / / api.certifiedcontent.org / data" is used as the public API.

[1506] 2. Authenticity verification: Verify the authenticity of the retrieved content by verifying its digital signature or authentication certificate.

[1507] 3. Data storage and formatting: Once content is verified as reliable, it is stored on a storage device, categorized by format and content, and redundant and duplicate data is removed. The organized dataset is then given metadata to enable efficient searching and learning.

[1508] 4. Training a generative AI model: A generative AI model is trained using the shaped dataset, using a framework such as TensorFlow or PyTorch, and is retrained every time new data is added.

[1509] Terminal

[1510] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays it to the user. The response is adjusted by the emotion engine and is delivered in a tone and expression that matches the user's emotions.

[1511] User

[1512] Users access the system through a terminal. They input questions or requests for information into the terminal, and the terminal sends the request to the server. The server's response is displayed on the terminal. The user can receive responses tailored by the emotion engine, resulting in a more satisfying experience.

[1513] Specific examples

[1514] Example 1: Nutrition Question

[1515] User Input

[1516] The user inputs the question "What foods are rich in vitamin C?" into the terminal. It is assumed that the user is slightly tired and in a negative emotional state.

[1517] Server Processing

[1518] The server receives the question and uses natural language processing technology to extract keywords such as "vitamin C" and "food." The emotion engine analyzes the user's emotions and recognizes negative emotions. The server then searches the database for relevant and reliable information, and the generative AI model generates an appropriate answer.

[1519] answer

[1520] The server generates a response such as the following and sends it to the device: "Foods that are rich in vitamin C include oranges, kiwis, and bell peppers. Eating these foods will help you stay healthy!" The response has an encouraging tone and is displayed on the device to the user.

[1521] This invention allows users to obtain highly reliable information in a tone and expression that suits their emotions at the time, preventing the spread of fake news and rumors and achieving high user satisfaction.

[1522] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1523] Step 1:

[1524] The server collects content whose authenticity has been confirmed from a certification authority. The input is data obtained through a public API or secure data transfer method. Specifically, the server uses the public API to call "https: / / api.certifiedcontent.org / data" and receives data in JSON format. The output is the raw data before its authenticity is confirmed.

[1525] Step 2:

[1526] The server verifies the digital signatures and authentication certificates of the collected content. The input is the content collected in step 1. Specifically, it verifies its authenticity using an algorithm that verifies the digital signature. The output is the data whose authenticity has been confirmed.

[1527] Step 3:

[1528] The server saves the content whose authenticity has been confirmed in the storage device. The input is the data whose authenticity has been confirmed in step 2. Specifically, it executes the "save" command to store the data in the database. The output is the trusted information as the data stored in the database.

[1529] Step 4:

[1530] The server analyzes the stored content and classifies it according to format and content. The input is the data in the database stored in step 3. It uses queries to classify the data into categories such as "nutrition," "health," and "emotions," and then refines the dataset by removing unnecessary and duplicate data. The output is the refined dataset.

[1531] Step 5:

[1532] The server trains the generative AI model using the formatted dataset. The input is the formatted dataset from step 4. Specifically, it runs a Python script and trains the model using TensorFlow or PyTorch. If necessary, it also retrains the model when new data is added. The output is a trained generative AI model.

[1533] Step 6:

[1534] The user inputs a question or request through the terminal. The input is text data representing the user's question or request. The terminal sends this request to the server. Specific operations include text input and transmission functions through the terminal's UI. The output is the request data sent to the server.

[1535] Step 7:

[1536] The server receives a request from the user and analyzes it. The input is the request data sent in step 6. Specifically, it uses a natural language processing library such as "NLTK" or "spaCy" to extract the intent and keywords of the question. The output is the analyzed intent and keyword information of the question.

[1537] Step 8:

[1538] The server runs an emotion engine that analyzes the emotion of the user's input text. The input is the question intent and keyword information extracted in step 7. The emotion analysis module is used to classify the emotion into positive, negative, and neutral categories. The output is data indicating the user's emotional state.

[1539] Step 9:

[1540] The server searches for relevant content in the database based on the analyzed request and emotional information. The input is the emotional state and question intent / keyword information obtained in step 8. It executes an SQL query to retrieve relevant information. The output is reliable relevant information.

[1541] Step 10:

[1542] The server uses the generative AI model to generate an answer based on the search results and sentiment analysis results. The input is the relevant information and emotional state obtained in step 9. The generative AI model inputs a prompt sentence to generate an appropriate answer. As a specific example, a "positive prompt sentence" is used to generate an answer with a positive tone. The output is the generated answer.

[1543] Step 11:

[1544] The server quality checks the generated answer to ensure there are no problems with grammar or content. The input is the answer generated in step 10. A quality check module is run to check the answer for accuracy and appropriateness. The output is the final answer that has passed the quality check.

[1545] Step 12:

[1546] The server sends the final answer to the device, which displays the answer to the user. The input is the final answer that passed the quality check in step 11. The server uses the Send API to send the answer to the device and updates the device's UI elements to display it to the user. The output is the answer displayed on the device in a form that the user can see.

[1547] (Application example 2)

[1548] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1549] Conventional generative AI systems have not adequately considered the reliability of content and user sentiment, resulting in a poor user experience and inaccurate information provided. Furthermore, the use of unreliable information can lead to problems such as the provision of inaccurate information and the spread of fake news.

[1550] 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.

[1551] In this invention, the server includes means for collecting content whose reliability has been confirmed by a certification authority, means for verifying the digital signatures and certification certificates of the collected content, means for storing the content whose reliability has been confirmed in a database, means for training a generative AI model based on the content in the database, means for recognizing and analyzing a user's emotions, means for adjusting the generated answers based on the analyzed emotions, means for quality checking the generated answers and checking for problems with grammar and content, and means for using a public API and secure data transfer means. This makes it possible to provide highly reliable information and optimal information tailored to the user's emotions.

[1552] A "generative AI model" is an algorithm that learns from collected data and generates new information and answers.

[1553] A "certification body" is an organization whose role is to verify and certify the reliability and accuracy of content.

[1554] "Confirmed reliable content" is information whose accuracy and integrity have been confirmed by a certification body.

[1555] A "digital signature" is an electronic means of verifying the origin of content or data and ensuring that it has not been tampered with.

[1556] "Certificate of Accreditation" means official evidence of an accreditation body's confirmation of trustworthiness.

[1557] An "emotion engine" is software that analyzes and classifies emotions from user input text and voice.

[1558] A "database" is a system that organizes and stores collected content and data.

[1559] "Quality check" is the process of checking the generated content for grammar and content appropriateness.

[1560] A "public API" is an interface designed to allow external access to specific functions or data.

[1561] "Secure data transfer methods" are technologies and protocols that prevent unauthorized access or tampering during data transmission.

[1562] A "prompt" is text that instructs a generative AI model on the format and content of input data.

[1563] This invention is a system that combines a generative AI system that uses content whose reliability has been confirmed by a certification organization with an emotion engine that recognizes user emotions. Specific embodiments of this system are described below.

[1564] Setup and Basic Operation

[1565] The main components of the system are the server, the terminals, and the users, each of which plays a specific role.

[1566] server

[1567] The server is responsible for collecting, validating, storing, learning, generating, and distributing data. Specifically, it works as follows:

[1568] 1. Content collection and verification: The server collects content that has been verified as trustworthy from a trusted authority, using public APIs and secure data transfer methods. The collected content is digitally signed and authenticated, and the server verifies it to confirm its authenticity.

[1569] 2. Storage and classification in a database: Once the content has been verified as reliable, it is stored in a database. The stored content is analyzed and classified according to its format and content. Unnecessary and duplicate data is removed, and the data is organized into a dataset.

[1570] 3. Training the generative AI model: The generative AI model is trained using the shaped dataset. The generative AI model learns patterns and relationships from large amounts of data and has the ability to generate new information and answers. As new data is added, the model is retrained, improving its accuracy.

[1571] 4. Request analysis and answer generation: Upon receiving a request from a user, the server analyzes the request. Natural language processing technology is used to extract the intent and keywords of the question. An emotion engine is also used to analyze the sentiment of the user's input text and classify it into positive, negative, and neutral sentiment categories. Based on this, the generative AI model generates an answer, and a quality check is performed to determine the final answer.

[1572] Terminal

[1573] The device acts as a user interface. When the user inputs a question or request, the device sends it to the server. When the device receives a response from the server, it displays the response to the user. The response is adjusted by the emotion engine and delivered in a tone and expression that matches the user's emotions.

[1574] User

[1575] Users access the system through a terminal. They input questions or requests for information into the terminal, and the answers from the server are displayed on the terminal. The user can receive answers tailored by the emotion engine, resulting in a more satisfying experience.

[1576] Specific examples

[1577] Suppose a user puts on smart glasses and inputs, "I'd like to know how to relax after a long day at work." This input is sent to the server via the device. The server analyzes the input, and its emotion engine classifies the user's emotional state as "fatigue." As a result, the server searches for appropriate product information (e.g., aroma diffuser or relaxation massager) from a trusted product database, and generates the following suggestion using a generative AI model: "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or relaxation massager. This aroma diffuser, in particular, has a lavender scent that is expected to promote sound sleep."

[1578] Prompt Sentence Examples

[1579] 1. Prompt to analyze emotions from user input:

[1580] User input text: "I want to know how to relax after a long day at work." Analyze the user's emotional state from this text.

[1581] 2. Prompt to generate sentiment-based product suggestions:

[1582] The user's emotional state has been analyzed as "fatigue." Please generate a sentence to explain to the user about a reliable aroma diffuser or relaxing massager.

[1583] In this way, users receive reliable information in a manner appropriate to their emotional state.

[1584] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1585] Step 1:

[1586] The device receives a user's question or request. The device receives input data by inputting a specific question or request in voice or text format. An example input might be a request such as, "I'd like to know how to relax after a long day at work." The input data is then sent directly to the next step.

[1587] Step 2:

[1588] The device sends the received user question or request to the server, using an Internet connection to transmit the user's input data to the server, which then receives the requested data.

[1589] Step 3:

[1590] The server analyzes the received request. Natural language processing technology is used to extract the intent of the question and key keywords from the text. For example, keywords such as "how to relax" and "long hours at work" may be extracted. Based on the results of this analysis, the system proceeds to the next processing step.

[1591] Step 4:

[1592] The server uses the emotion engine to analyze the emotion of the user's input text. For example, if the user's input is "I want to know how to relax after a long day of work," the emotion engine will classify the user's emotional state as "fatigue" from this text. The analysis result is saved as an emotion category.

[1593] Step 5:

[1594] The server searches for relevant information from a trusted content database based on the analyzed emotional state and key keywords. For example, it searches for content related to "how to relax" or "fatigue" and retrieves the results. These results are then used.

[1595] Step 6:

[1596] The server generates an answer based on the search results using a generative AI model. The generative AI model uses patterns and relationships learned from large amounts of data to generate an appropriate answer. For example, a sentence such as "Thank you for your long hours at work. To help you relax, we recommend an aroma diffuser or a relaxing massage machine" is generated. This generated sentence is then used.

[1597] Step 7:

[1598] The server then performs a quality check on the generated answer, checking for grammar and content issues and making corrections as necessary. Once the answer passes this check, it is used.

[1599] Step 8:

[1600] The server then adjusts the generated answer based on the results of the emotion analysis. To match the tired emotion, the server may change the tone to something gentler and more relaxing. The final answer is then sent in the next step.

[1601] Step 9:

[1602] The server sends the final answer to the device, using an Internet connection to transmit the answer data to the device, which can then receive the data from the server.

[1603] Step 10:

[1604] The device then displays the received response to the user. The response may be in the form of text or voice, and is delivered in an appropriate tone and expression. For example, a message such as "Thank you for your long hours at work. To help you relax, we recommend using an aroma diffuser or a relaxing massager" may be displayed.

[1605] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice 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 voice data.

[1606] 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> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) 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.

[1607] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1608] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1609] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1610] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1611] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1612] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1613] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1614] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1615] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1616] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1617] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1618] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1619] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1620] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1621] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1622] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1623] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1624] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1625] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1626] The following is further disclosed regarding the above embodiment.

[1627] (Claim 1)

[1628] A means of collecting content whose authenticity has been verified by a certification body;

[1629] A means to verify digital signatures or authentication certificates on collected content;

[1630] a means for storing the verified authentic content in a database;

[1631] A means to train a generative AI model based on the content in the database, and

[1632] means for receiving a request from a user and analyzing the request;

[1633] means for searching for relevant content in a database based on the parsed request;

[1634] A system that includes a means for a generative AI model to generate answers based on search results.

[1635] (Claim 2)

[1636] A means to analyze the collected content, classify it according to format and content, remove unnecessary and duplicate data, and format it into a dataset;

[1637] 10. The system of claim 1, further comprising means for quality checking the generated answers to ensure grammatical and content integrity.

[1638] (Claim 3)

[1639] 10. The system of claim 1, further comprising means for using a public API and secure data transfer means to periodically obtain certified content from a certification authority.

[1640] "Example 1"

[1641] (Claim 1)

[1642] A means of collecting content whose authenticity has been verified by a certification body;

[1643] A means to verify digital signatures or authentication certificates on collected content;

[1644] a means for storing the verified authentic content in a database;

[1645] A means to analyze the stored content, classify it according to format and content, remove unnecessary and duplicate data, and format it into a dataset;

[1646] A means to train a generative AI model based on the formatted dataset in the database; and

[1647] means for receiving a request from a user and analyzing the request;

[1648] means for searching for relevant content in a database based on the parsed request;

[1649] A means for the AI ​​model to generate answers based on the search results, and

[1650] A system that includes a means to quality check the generated answers to ensure they are grammatically and content-wise correct.

[1651] (Claim 2)

[1652] 10. The system of claim 1, further comprising means for using a public API and secure data transfer means to periodically obtain certified content from a certification authority.

[1653] (Claim 3)

[1654] 2. The system according to claim 1, further comprising means for analyzing a request from a user using natural language processing technology and extracting intent and keywords.

[1655] "Application Example 1"

[1656] (Claim 1)

[1657] A means of collecting content whose authenticity has been verified by a certification body;

[1658] A means to verify digital signatures or authentication certificates on collected content;

[1659] a means for storing the verified authentic content in a database;

[1660] A means to train a generative AI model based on the content in the database, and

[1661] means for receiving a request from a user and analyzing the request;

[1662] means for searching for relevant content in a database based on the parsed request;

[1663] A means for the AI ​​model to generate answers based on the search results, and

[1664] A means to assess the authenticity of web pages and social media content that users are viewing in real time;

[1665] and means for displaying a warning for untrustworthy content.

[1666] (Claim 2)

[1667] A means to analyze the collected content, classify it according to format and content, remove unnecessary and duplicate data, and format it into a dataset;

[1668] A means to quality check the generated answers to ensure there are no problems with grammar or content, and

[1669] 10. The system of claim 1, further comprising: means for displaying a warning to a user about untrustworthy content.

[1670] (Claim 3)

[1671] A means to periodically retrieve certified content from certification bodies using public APIs and secure data transfer methods;

[1672] 2. The system according to claim 1, further comprising means for providing a real-time warning to a user about the authenticity of the content being viewed based on the result of the authenticity evaluation.

[1673] "Example 2: Combining Emotion Engines"

[1674] (Claim 1)

[1675] A means of collecting content whose authenticity has been verified by a certification body;

[1676] A means to verify digital signatures or authentication certificates on collected content;

[1677] means for storing the content whose authenticity has been confirmed in a storage device;

[1678] a means for training a generative AI model based on the content in the storage device;

[1679] means for receiving a request from a user and analyzing the request;

[1680] means for searching for relevant content in a storage device based on the parsed request;

[1681] A means for adjusting the content and tone of responses based on the results of an emotion engine that analyzes the user's emotions;

[1682] A system that includes a means for a generative AI model to generate answers based on search results and sentiment analysis results.

[1683] (Claim 2)

[1684] A means to analyze the collected content, classify it according to format and content, remove unnecessary and duplicate data, and format it into a dataset;

[1685] 10. The system of claim 1, further comprising means for quality checking the generated answers to ensure grammatical and content integrity.

[1686] (Claim 3)

[1687] 10. The system of claim 1, further comprising means for using a public API and secure data transfer means to periodically obtain certified content from a certification authority.

[1688] "Application example 2 when combining emotion engines"

[1689] (Claim 1)

[1690] A means of collecting content whose authenticity has been verified by a certification body;

[1691] A means to verify digital signatures or authentication certificates on collected content;

[1692] a means for storing the verified authentic content in a database;

[1693] A means to train a generative AI model based on the content in the database, and

[1694] means for receiving a request from a user and analyzing the request;

[1695] means for searching for relevant content in a database based on the parsed request;

[1696] A means for the AI ​​model to generate answers based on the search results, and

[1697] means for recognizing and analyzing user emotions;

[1698] The system includes means for adjusting the generated answers based on the analyzed sentiment.

[1699] (Claim 2)

[1700] A means to analyze the collected content, classify it according to format and content, remove unnecessary and duplicate data, and format it into a dataset;

[1701] A means to quality check the generated answers to ensure there are no problems with grammar or content, and

[1702] 10. The system of claim 1, further comprising means for generating suggestions based on the sentiment analysis results and providing the suggestions in an appropriate tone and expression.

[1703] (Claim 3)

[1704] 10. The system of claim 1, further comprising means for using a public API and secure data transfer means to periodically obtain certified content from a certification authority. [Explanation of symbols]

[1705] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting content whose authenticity has been verified by a certification body; A means to verify digital signatures or authentication certificates on collected content; a means for storing the verified authentic content in a database; A means to train a generative AI model based on the content in the database, and means for receiving a request from a user and analyzing the request; means for searching for relevant content in a database based on the parsed request; and a means for a generative AI model to generate an answer based on the search results.

2. A means to analyze the collected content, classify it according to format and content, remove unnecessary and duplicate data, and format it into a dataset; 10. The system of claim 1, further comprising means for quality checking the generated answers to ensure grammatical and content integrity.

3. 10. The system of claim 1, further comprising means for using a public API and secure data transfer means to periodically obtain certified content from a certification authority.

Images