system

Abstract

Provide a system. A method for a user to verify the authenticity of generated text, a means for displaying the generated text; means for providing a button to request a fact check on the text; means for transmitting the text to a server when a fact-check request is made by a user; means for adding received fact-check requests to a queue; A means for a collaborator to obtain a fact-check request from the queue, check the content, and send the result to the server; means for notifying a user of the fact-check results received by the server; A means for feeding back the results as learning data for the AI; A system including:

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] In recent years, text generation technology, including generative AI, has rapidly developed, allowing people to easily access a wide variety of information. However, there is no guarantee that the content of the generated text is necessarily true, and it may contain misinformation or inaccuracies. In such situations, users want to fact-check the generated text to verify its authenticity, but each fact-check requires significant resources and knowledge, placing a heavy burden on users. Traditional methods require manual verification and cross-referencing with other sources, which is time-consuming and labor-intensive. This leads to the spread of misinformation and a decline in credibility. [Means for solving the problem]

[0005] To solve the above-mentioned problems, the present invention provides a system that allows users to easily verify the authenticity of generated text. Specifically, it provides a means for displaying the generated text and a button for requesting a fact-check of the text. When the user presses this button, the text is sent to a server, and the server adds the received fact-check request to a queue. Contributors retrieve requests from this queue, check the content, and send the results to the server. The server notifies the user of the received fact-check results and also feeds the results back as learning data for the AI. In addition, it provides a means for rewarding contributors who provide excellent fact-check answers, encouraging active participation from contributors. This allows users to obtain reliable information quickly and efficiently, preventing the spread of misinformation.

[0006] "User" refers to a person or organization that uses a generative AI system to generate text and request fact-checking.

[0007] "Generated text" refers to documents or content that are output and displayed by generative AI.

[0008] "Fact-checking" refers to the process of verifying and evaluating the factuality of the content of generated text.

[0009] "Server" refers to the part of the system that receives fact-checking requests from users, sends the requests to contributors, and stores and notifies the results.

[0010] A "collaborator" refers to a person or organization whose role is to receive fact-checking requests from the server, verify the content, and return the results to the server.

[0011] "Queue" refers to a data structure used to process multiple fact-check requests in sequence.

[0012] "Training data" refers to fact-checked data used to improve and enhance AI models.

[0013] "Rewards" refers to the compensation or incentives given to contributors for conducting fact-checks and providing excellent results. [Brief explanation of the drawings]

[0014] [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 showing 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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0036] First, a user accesses the generative AI system using their own device and generates text. The generated text is immediately displayed on the user's device, with a button next to the text to request fact-checking.

[0037] When a user presses this button, the corresponding text is sent to the server. The server adds the received text to a queue as a fact-check request. After the fact-check request is added to the queue, a contributor retrieves the request from the queue and works to verify the authenticity of the text. The contributor then sends the verification result to the server, which receives and stores the result.

[0038] The server then notifies the user of the received fact-check results. Specifically, the results are sent to the user's device and displayed on the user interface. The fact-check results are also used as training data for the AI ​​model. The server feeds the results back into the AI ​​model to improve its accuracy and reliability.

[0039] Furthermore, contributors who provide excellent fact-checking results will be rewarded. This reward system is expected to increase contributors' motivation and lead to higher quality fact-checking.

[0040] As a specific example, consider the case where a user generates text such as "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to a queue. A collaborator retrieves this request from the queue and obtains the check result, "Mount Everest is the highest." The result is sent to the server, which notifies the user of the result and provides feedback to the AI ​​model.

[0041] Each process in this system is carried out as follows: The user checks the text generated by the generative AI on their own device and sends a request to the server via the fact-check button. The server adds the request to a queue, receives and stores the results from contributors, and notifies the user. Finally, feedback is provided to the AI ​​model and contributors are rewarded. This series of steps allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole.

[0042] The above is an embodiment of the present invention. This system allows users to verify the authenticity of text without any burden, and enables them to quickly obtain accurate information.

[0043] The processing flow will be explained below.

[0044] Step 1:

[0045] A user uses a generative AI system to request text generation. Based on this request, the server instructs the generative AI to generate text. The generated text is sent from the server to the user's device and displayed on the device's screen.

[0046] Step 2:

[0047] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the generated text.

[0048] Step 3:

[0049] The user clicks the "Outsource fact-checking" button. The device takes the generated text and sends a fact-checking request in JSON format to the server.

[0050] Step 4:

[0051] The server processes the received fact-check request by adding it to a queue and notifying the user that the request has been received.

[0052] Step 5:

[0053] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor uses their own resources and knowledge to verify the authenticity of the generated text.

[0054] Step 6:

[0055] The collaborators compile the verification results and send them to the server, which receives them and stores them in a database.

[0056] Step 7:

[0057] The server notifies the user of the saved fact-check results, which are then displayed on the user's device, allowing the user to view the results.

[0058] Step 8:

[0059] The server feeds the fact-checking results back into the AI ​​model, which then undergoes a learning process to improve the model's accuracy and reliability.

[0060] Step 9:

[0061] The server evaluates the quality of the fact-check results provided by contributors and rewards contributors who provide excellent results. The server manages the distribution of rewards and notifies contributors.

[0062] This allows users to efficiently verify the authenticity of generated text, improving the reliability of the entire system.

[0063] Example 1

[0064] 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."

[0065] Currently, many generative AI systems generate text for users, but the lack of a means to verify the authenticity of the generated text has led to the spread of misinformation. Furthermore, the fact-checking process is often done manually, which is time-consuming and the quality of the checks cannot be guaranteed. Furthermore, there is a lack of mechanisms to maintain the motivation of contributors and continuously conduct high-quality fact-checks. Therefore, there is a need for the development of a system that can quickly and accurately verify the authenticity of generated text.

[0066] 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.

[0067] In this invention, the server includes: means for displaying the generated text so that the user can verify the authenticity of the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a contributor to obtain the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the artificial intelligence; means for displaying the result on the user's terminal based on the generated text; and means for calculating and notifying a reward to the contributor. This allows the user to quickly and accurately verify the authenticity of the generated text, improving the reliability of the entire system.

[0068] A "user terminal" is a device used by a user to display the generated text and request fact-checking.

[0069] "Generative AI" is artificial intelligence that generates text in response to a prompt.

[0070] The "fact check button" is an operation unit that allows the user to request confirmation of the authenticity of the generated text.

[0071] The "server" is a device that receives fact-checking requests, adds them to a queue, notifies the user of the results, and provides feedback to the artificial intelligence.

[0072] A "queue" is a data structure that allows a server to manage fact-check requests sequentially.

[0073] A "collaborator" is an individual or organization that responds to a fact-check request and verifies the authenticity of the text.

[0074] "Fact-check results" are information about the authenticity of a text that has been confirmed by a contributor.

[0075] "Artificial intelligence training data" refers to fact-checking results data used to improve the accuracy and reliability of generative AI.

[0076] "Rewards" are compensation paid to contributors who provide excellent fact-checking results.

[0077] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0078] First, a user accesses the generative AI system using their own device and generates text. Specifically, they access the generative AI system's interface using a web browser. The user inputs a prompt sentence. For example, they input the prompt sentence, "Mount Fuji is the highest mountain in the world." The generative AI model (e.g., OpenAI's GPT-3 (registered trademark)) generates text based on this prompt sentence. The generated text is immediately displayed on the user's device.

[0079] Next to the generated text is a button for requesting a fact check. When a user clicks this fact check button, the text is sent to the server as a fact check request. The server processes the received fact check request and adds it to a queue. This queue is a data structure for managing the processing order of requests.

[0080] The collaborator retrieves a request from the queue using a dedicated interface. The collaborator verifies the authenticity of the text content, for example, using the Internet or literature. For example, for the text "Mount Fuji is the highest mountain in the world," the collaborator obtains the result "Mount Everest is the highest." The verification result is entered into a dedicated input form and sent to the server.

[0081] The server stores the received fact-check results and notifies the user. The results are displayed on the user's device. For example, the result "Everest is the highest mountain" is displayed. The server also stores the fact-check results in a database for feedback to the generative AI model. This feedback improves the accuracy and reliability of the generative AI model.

[0082] In addition, contributors who provide excellent fact-checking results are rewarded. This reward can be paid via PayPal, for example. The server calculates the reward for contributors and notifies them.

[0083] As a concrete example of this system, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to the queue. A collaborator retrieves this request from the queue and obtains the check result "Mount Everest is the highest." The result is sent to the server, which notifies the user and provides feedback to the generative AI model.

[0084] This process allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole. Furthermore, by providing appropriate rewards to contributors, it is expected that high-quality fact-checking will be carried out sustainably. This concludes the description of the present invention.

[0085] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0086] Step 1:

[0087] A user accesses a generative AI system and inputs a prompt sentence. For example, they access the system's interface using a web browser. They input the prompt sentence, "Mount Fuji is the highest mountain in the world." Based on this input, a generative AI model (e.g., OpenAI's GPT-3) generates text. The generated text is displayed on the user's device screen. The input here is the prompt sentence, and the output is the generated text.

[0088] Step 2:

[0089] The user clicks on the fact-check button located next to the generated text. This click becomes the input, and a fact-check request is sent to the server along with the generated text. The server processes the received text as a fact-check request and adds it to a queue, which is a data structure. Here, the input is the user's click and the generated text, and the output is the server adding the fact-check request to the queue.

[0090] Step 3:

[0091] The server adds fact-check requests to a queue and monitors their status. Contributors retrieve requests from the queue using a dedicated interface. The retrieved text is sent to the contributor's device, and the contributor begins work to verify the text's content. The input here is the fact-check request added to the queue, and the output is the text received on the contributor's device.

[0092] Step 4:

[0093] A collaborator verifies the authenticity of the text content using the Internet or literature. For example, for a text that reads "Mount Fuji is the highest mountain in the world," the verification result is "Mount Everest is the highest mountain." The collaborator enters the verification result into a dedicated input form and sends it to the server. The input here is the text content and verification materials, and the output is the verification result.

[0094] Step 5:

[0095] The server stores the received confirmation result. The server notifies the user's terminal of this result. For example, the result "Everest is the highest mountain" is displayed on the user's terminal. The input here is the confirmation result, and the output is the notification to the user terminal.

[0096] Step 6:

[0097] The server feeds back the fact-check results to the generative AI model. The received results are stored in a database and used as training data for the generative AI model. The input here is the stored fact-check results, and the output is a reflection on the training data for the generative AI model.

[0098] Step 7:

[0099] The server calculates rewards for contributors. Contributors who provide good fact-check results are paid, for example, via PayPal. The payment of rewards includes a notification from the server to the contributor. The input here is the evaluation of the fact-check result, and the output is the payment of rewards to contributors.

[0100] (Application example 1)

[0101] 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."

[0102] The challenge is to quickly verify the reliability of text and product information generated by generative AI models, ensuring that users can safely use accurate information. It is also necessary to ensure the authenticity of product information provided in virtual stores, thereby improving users' shopping experiences.

[0103] 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.

[0104] In this invention, the server includes: means for a user to verify the authenticity of generated text; means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for an AI; means for providing product information to the user in a virtual store; means for displaying a button in the virtual store for requesting a fact-check on the generated product information; and means for requesting a fact-check through the product information displayed on the user's smart device. This allows users to quickly verify the authenticity of product information provided in the virtual store and enjoy shopping with peace of mind.

[0105] "User" refers to a person who uses a generative AI model to generate text or product information and verify the authenticity of that content.

[0106] "Generated text" refers to sentences or descriptions automatically created by a generative AI model.

[0107] "Server" refers to the computer system that manages the generated text and fact-checking requests, processes the fact-checking results from contributors, and notifies the user.

[0108] "Fact checking" refers to the process of verifying the authenticity of the content of generated text or product information.

[0109] A "collaborator" is someone who receives a fact-check request sent from the server, checks its contents, and sends the results to the server.

[0110] A "queue" refers to a data structure that stores received fact-check requests in order and provides them to contributors.

[0111] "AI training data" refers to data used to improve the accuracy and reliability of generative AI models.

[0112] "Virtual store" refers to a virtual store environment where products are sold and purchased over the Internet.

[0113] "Product information" refers to information such as product features, specifications, reviews, etc. presented in the virtual store.

[0114] "Smart devices" refers to portable electronic devices that can connect to the Internet, such as smartphones and smart glasses.

[0115] The present invention provides a system for outsourcing fact-checking to verify the authenticity of text and product information output by a generative AI model. The following describes in detail an embodiment of the present invention.

[0116] System Overview

[0117] The system consists of the following elements:

[0118] 1. User Device

[0119] 2. Cloud Server

[0120] 3. Collaborator terminal

[0121] 4. Generative AI Models

[0122] 5. Smart devices (smartphones, smart glasses)

[0123] System Operation

[0124] User device:

[0125] The user terminal provides a user interface for generating text and product information. The user inputs a prompt sentence into the generative AI model to generate the text and product information. The generated text and product information are displayed on the screen of the user terminal.

[0126] Cloud Server:

[0127] The cloud server plays a central role in managing fact-checking requests. When a user submits a fact-checking request, it is sent to the server and added to a queue. The server then provides the fact-checking requests to contributors and collects the results.

[0128] Collaborator's device:

[0129] The contributor terminal provides an interface for receiving fact-check requests. The contributor processes the requests and sends the results to the cloud server. Contributors are rewarded for providing good fact-check answers.

[0130] Generative AI models:

[0131] A generative AI model (e.g., GPT-4 (registered trademark)) generates text and product information based on prompts entered by users. This model runs on a cloud server and provides information in response to requests from user devices.

[0132] Smart Devices:

[0133] Users use smart devices (e.g., smartphones or smart glasses) to view the generated text and product information and submit fact-check requests.

[0134] Data processing and calculation

[0135] The cloud server processes data using the following software and hardware:

[0136] software:

[0137] Generative AI models (e.g., GPT-4)

[0138] Virtual store platforms (e.g., Shopify, WooCommerce)

[0139] Task management systems (e.g., Trello)

[0140] Hardware:

[0141] Cloud-based servers (e.g., Amazon Web Services) that provide computing power

[0142] Detailed procedure

[0143] 1. Information generation and display:

[0144] When a user accesses a product page in the virtual store, the generative AI model generates product information and displays it on the user's smart device.

[0145] 2. Fact-checking requests:

[0146] When a user presses the fact-check button, the generated text and product information are sent to the cloud server and added to the queue as a fact-check request.

[0147] 3. Collaborator verification:

[0148] The collaborator retrieves the request from the queue and checks the authenticity of the product information. The result is sent to the cloud server and notified to the user.

[0149] 4. Feedback of results:

[0150] The cloud server feeds the received results back into the AI ​​model to improve the model's accuracy.

[0151] Specific examples

[0152] For example, imagine a user wearing smart glasses accesses a virtual store and browses the product page for the latest smartphone. At this time, the generative AI model generates information such as "The latest smartphones are 5G compatible, have high-resolution cameras, and have long-lasting batteries." If the user wants to verify the authenticity of this information, they can click the fact-check button. The results of the check by the collaborators are then notified, and the user can receive reliable information such as "This smartphone is actually 5G compatible and has a long-lasting battery."

[0153] Prompt Sentence Examples

[0154] Generate user-specified product features and specifications.

[0155] Explain the effect of the fact-check button on product information.

[0156] This system allows users to quickly verify the authenticity of product information provided in virtual stores, allowing them to enjoy shopping with peace of mind.

[0157] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0158] Step 1:

[0159] The user device sends a request to the generative AI model.

[0160] Specifically, the user enters a prompt sentence and clicks the generate button. This sends the prompt sentence to the cloud server as an API request. The input data is the prompt sentence, and the output data is the generated text. The server invokes the generative AI model, generates text based on the entered prompt sentence, and sends the result to the user's device.

[0161] Step 2:

[0162] The user terminal displays the generated text.

[0163] The generated text received from the server is displayed on the user terminal. The user checks this text and determines its authenticity. The input data is the generated text received from the server, and the output data is the text displayed on the user interface.

[0164] Step 3:

[0165] The user terminal sends a fact-check request to the server.

[0166] When a user clicks the fact-check button, the generated text is sent to the cloud server as a fact-check request. The input data is the generated text and the fact-check request action, and the output data is the data to be added to the queue as a fact-check request. The server adds the request to a queue data structure.

[0167] Step 4:

[0168] The server provides fact-check requests to contributors.

[0169] The server retrieves fact-check requests from the queue and provides them to the collaborator terminal. The input data is the fact-check request in the queue, and the output data is the request sent to the collaborator terminal. The collaborator receives the request and starts checking it.

[0170] Step 5:

[0171] Contributors submit fact-check results to the server.

[0172] The contributor processes the request and sends the results to the cloud server. The input data is the fact-check request, and the output data is the verification result. The server stores the results received from the contributor and prepares them for further processing.

[0173] Step 6:

[0174] The server notifies the user terminal of the fact-check results.

[0175] The server sends the received fact-check results to the user's device. The input data is the confirmation result received from the collaborator, and the output data is a notification displayed on the user's device. The user can check the result.

[0176] Step 7:

[0177] The server feeds the fact-checking results back to the generative AI model.

[0178] The server feeds back the fact-check results as training data for the generative AI model. This feedback improves the accuracy and reliability of the AI ​​model. The input data is the fact-check results, and the output data is an updated AI model.

[0179] Step 8:

[0180] The user's smart device displays the product information.

[0181] The generated product information is displayed on the smart device, allowing for fact-checking requests. The input data is the generated product information, and the output data is the information displayed on the smart device. The user can click to submit the request.

[0182] 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.

[0183] The present invention is a system that outsources fact-checking to verify the authenticity of the content of text output by generative AI, and further combines it with an emotion engine that recognizes the emotional state of the user and controls the operation of the system based on that. Below, we will explain in detail the embodiments of the present invention.

[0184] First, the user accesses the generative AI system using their own device and generates text. This generated text is immediately displayed on the user's device. Next to the generated text, there is a button to request a fact check. This button is used if the user wants to verify the authenticity of the generated text.

[0185] Next, when the user presses this button, the corresponding text is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received text and emotional state information to a queue as a fact-check request and notifies the user of the receipt of the request.

[0186] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated text and sends the results to the server. The server receives the results and stores them in a database. The saved fact-check results are notified to the user and displayed on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[0187] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[0188] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[0189] As a specific example, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check result, and the server notifies the user of the result. At the same time, the result and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[0190] The above is an embodiment of the present invention. The introduction of an emotion engine enables flexible fact-checking that takes into account the emotional state of the user, further improving the usability and reliability of the system.

[0191] The processing flow will be explained below.

[0192] This invention is a system that outsources fact-checking to verify the authenticity of the text content output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that. The processing flow is explained below in specific steps.

[0193] Step 1:

[0194] A user accesses the generative AI system using a device and requests text generation. The server instructs the generative AI to generate text and sends the generated text to the user's device. The device then displays the generated text on its screen.

[0195] Step 2:

[0196] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the text.

[0197] Step 3:

[0198] The user clicks the "Outsource fact-checking" button. The device retrieves the generated text, and the emotion engine recognizes the user's current emotional state. This information is then sent to the server in JSON format as a fact-checking request.

[0199] Step 4:

[0200] The server adds the received fact-check request to a queue and notifies the user with a message confirming receipt of the request.

[0201] Step 5:

[0202] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor verifies the authenticity of the generated text and sends the result to the server.

[0203] Step 6:

[0204] The server receives fact-check results from contributors, stores them in a database, and notifies the user of the results and displays them on the user's device.

[0205] Step 7:

[0206] The server then feeds back the stored fact-check results and the user's emotional state information to the AI ​​model, which is used to improve the model's performance.

[0207] Step 8:

[0208] The emotion engine determines the need for fact-checking based on the user's emotional state. If the user expresses anxiety or doubt, the system automatically sends a fact-check request to the server, reducing the burden on the user.

[0209] Step 9:

[0210] The server evaluates the quality of the fact-check results provided by contributors, rewards contributors who provide excellent results, and manages the distribution of rewards and notifies contributors.

[0211] Through these steps, users can efficiently verify the authenticity of the generated text, improving the reliability and usability of the entire system. The introduction of an emotion engine enables flexible responses that take into account the user's psychological state, quickly ensuring the accuracy of information.

[0212] Example 2

[0213] 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."

[0214] There is a need for a method to quickly and accurately verify the authenticity of user-generated documents. Conventional systems have problems: fact-checking takes time and it is difficult to respond appropriately taking into account the user's emotional state. This can lead to users becoming anxious about the reliability of the information, which can lead to a decrease in the reliability of the system.

[0215] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for displaying a document created by a user, means for providing a button for requesting a fact-check for the document, means for transmitting the document to the server when a fact-check request is made by the user, means for adding the fact-check request received by the server to a queue, means for a collaborator to acquire the fact-check request from the queue, check the content, and transmit the result to the server, means for notifying the user of the fact-check result received by the server, means for feeding back the result and information on the user's emotional state as learning data for the AI, and an emotion engine that recognizes the user's emotional state and automatically requests a fact-check. This makes it possible to quickly and accurately verify the authenticity of a document while taking the user's emotional state into consideration.

[0216] A "user" is an entity that uses the system to generate documents and request fact-checking.

[0217] "Document" refers to text data created by generative AI.

[0218] "Generative AI" is an artificial intelligence model that generates documents based on prompts entered by the user.

[0219] "Fact checking" is a verification process to confirm the authenticity of the contents of a document.

[0220] A "button" is an interface element that a user clicks on the screen to request a specific action from the system.

[0221] A "server" is a computer system that processes and stores user requests and data.

[0222] A "queue" is a data structure for queuing received fact-check requests.

[0223] A "collaborator" is a person or system that receives a fact-check request, verifies its contents, and provides the results.

[0224] "Emotional state information" is data that represents the emotional state of the user.

[0225] "Training data" is a collection of past data and results used to improve the performance of an AI model.

[0226] An "emotion engine" is software or hardware that recognizes the user's emotional state and controls the system's behavior based on that.

[0227] "Feedback" refers to the re-input and re-processing of specific data or results for the purpose of improving the system.

[0228] "Notification" is a communication method by which a server provides information to a user.

[0229] A "prompt" refers to an instruction or question that is input to a generative AI.

[0230] This invention is a system that outsources fact-checking to verify the authenticity of the contents of documents output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that.

[0231] First, a user uses their own device to access a generative AI system (e.g., a generative AI model) and enters a prompt to generate a document. The generated document is immediately displayed on the user's device. Next to the generated document, there is a button for requesting fact-checking. This button is used when the user wants to verify the authenticity of the generated document.

[0232] Next, when the user presses the fact-check button, the document is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received document and emotional state information to a queue (e.g., AWS (registered trademark) SQS) as a fact-check request and notifies the user of the receipt of the request.

[0233] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated document content and sends the results to the server. The server receives the results and stores them in a database (e.g., MySQL (registered trademark)). The server notifies the user of the stored fact-check results and displays them on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[0234] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[0235] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[0236] As a specific example, consider the case where a user generates a document stating "Mount Fuji is the tallest mountain in the world" and requests that it be fact-checked. This document and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check results, and the server notifies the user of the results. At the same time, the results and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[0237] The introduction of an emotion engine enables flexible fact-checking that takes into account the user's emotional state, further improving the usability and reliability of the system.

[0238] An example of a prompt sentence is "Generate the document 'Mount Fuji is the highest mountain in the world' and request a fact check."

[0239] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0240] Step 1:

[0241] A user accesses the generative AI system using their own device. The user inputs a prompt and requests text generation. The device sends the user's input (prompt) to the generative AI model and receives the generated text. For example, if a user inputs the prompt "Tell me about Mount Fuji," the device receives the text "Mount Fuji is the highest mountain in Japan" from the generative AI. This generated text is immediately displayed on the device.

[0242] Step 2:

[0243] The user presses a fact-check button displayed next to the generated text. The device sends the generated text and the user's emotional state (e.g., anxiety or doubt) to the server. The input here is the generated text and the user's emotional state information, which the device then sends to the server. For example, the device sends the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state information of anxiety.

[0244] Step 3:

[0245] The server adds the received text and emotional state information to a queue. As input, the server receives the generated text and emotional state information received from the device. The server adds this information to a queue (e.g., AWS SQS) and processes it as a fact-checking request. For example, the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state are added to the queue.

[0246] Step 4:

[0247] The server notifies the user that it has received the fact-checking request. The server confirms that the fact-checking request has been added to the queue and notifies the user that it has been received. The input is the queue information added to the server, and the output is a notification to the user. For example, a notification saying "your fact-checking request has been accepted" is sent to the user.

[0248] Step 5:

[0249] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. As input, the contributor retrieves a fact-check request from the queue, and as output, the retrieved text information is obtained. For example, a contributor retrieves the text "Mount Fuji is the highest mountain in Japan."

[0250] Step 6:

[0251] The collaborator verifies the truth of the generated text content and sends the result to the server. As input, the collaborator receives the text information to be verified and investigates its content. As output, the investigation result is sent to the server. For example, the collaborator verifies the result that "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" and sends it to the server.

[0252] Step 7:

[0253] The server receives the fact-check results and stores them in a database. As input, it receives the fact-check results sent by the contributors, and as output, it stores them in a database. For example, the result "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" is stored in the database.

[0254] Step 8:

[0255] The server notifies the user of the saved fact-check results and displays the results on the user's device. As input, the server retrieves the saved results from the database and notifies the user as output. The user can view the fact-check results on their device. For example, a notification is displayed to the user saying, "Fact-check result: Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world."

[0256] Step 9:

[0257] The server feeds back the fact-check results and the user's emotional state information as learning data for the AI ​​model. The fact-check results and emotional state information are used as input, and data to improve the accuracy of the AI ​​model is obtained as output. For example, the text "Mount Fuji is the highest mountain in Japan," the emotional state "anxiety," and the fact-check results are fed back to the AI ​​model.

[0258] Step 10:

[0259] The emotion engine recognizes the user's emotional state and automatically requests fact-checking when necessary. The user's emotional state information is used as input, and an automatic fact-checking request is generated as output. For example, if the user expresses anxiety, the system will automatically request a fact-check.

[0260] By using the above steps, the system of the present invention can quickly and accurately verify the authenticity of a document while taking into account the emotional state of the user.

[0261] (Application example 2)

[0262] 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."

[0263] In recent years, text generation using generative AI has become widespread, and the importance of fact-checking to verify the authenticity of generated text is increasing. However, the process for users to verify the reliability of generated text is time-consuming and often leaves users feeling anxious or suspicious. Furthermore, a system that takes into account the user's emotional state is needed to efficiently verify the authenticity of generated text. Furthermore, a reward system for contributors is also necessary to improve the quality of fact-checking results. The present invention aims to provide a system that solves these problems.

[0264] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a user requests a fact-check; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, check the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the AI; means for recognizing the user's emotional state using an emotion engine and controlling the system operation based on that information; means for automatically requesting a fact-check according to the emotional state; and means for displaying the fact-check result based on the emotional state on the user's terminal. This allows the user to quickly and efficiently verify the authenticity of the generated text, thereby realizing flexible fact-checking that takes the user's emotional state into consideration.

[0265] "User" means an individual or organization that uses the generative AI system.

[0266] "Generated text" refers to textual information generated by a generation AI.

[0267] "Fact checking" is the act of verifying the truth of the content of the generated text.

[0268] The "emotion engine" is an engine that recognizes the user's emotional state and reflects that information in the system's operation.

[0269] "Server" means the computer system that manages and processes data and fact-checking requests submitted by users.

[0270] A "collaborator" is an individual or organization whose role is to receive a fact-check request, verify its contents, and return the results.

[0271] A "queue" is a data structure for lining up received fact-check requests in order and waiting.

[0272] "Notification" means the act or means of communicating fact-check results or other information to a user.

[0273] "Training data" is data used to train and improve AI models.

[0274] The "reward system" is a mechanism that rewards contributors for providing excellent fact-checking results.

[0275] "Generative AI" is artificial intelligence that allows users to generate text.

[0276] A "terminal" is a device through which a user accesses the generative AI system.

[0277] The present invention provides a system for accessing a generation AI system on a user's terminal and verifying the authenticity of generated text. Specific embodiments for carrying out the present invention will be described below.

[0278] Overall system configuration

[0279] The system consists of the following main components:

[0280] 1. User devices: smartphones, tablets, computers, etc.

[0281] 2. Server: Manages the queue of fact-checking requests and stores the results.

[0282] 3. Generative AI: An AI engine that allows users to generate text.

[0283] 4. Emotion engine: Recognizes the user's emotional state and controls the system behavior.

[0284] 5. Fact-checking contributors: These contributors are responsible for verifying the authenticity of the generated text.

[0285] Program Overview

[0286] 1. Text Generation and Display

[0287] Users can use generative AI through their device to generate specific text, which is then instantly displayed on the device.

[0288] 2. Recognizing emotional states

[0289] The emotion engine uses the device's camera, microphone, and sensors to recognize the user's emotional state in real time, for example, using facial recognition and voice analysis technologies.

[0290] 3. Fact-checking requests

[0291] A button to request a fact check will be displayed next to the generated text. If the user has any concerns or doubts, they can press this button to send a fact check request to the server.

[0292] 4. Queue Management

[0293] The server adds the received fact-check request to a queue, and at the same time, the emotional state data sent from the emotion engine is also recorded on the server.

[0294] 5. Fact Check

[0295] Contributors access the server using a dedicated interface to retrieve fact-check requests added to the queue, check their contents, and send the results to the server.

[0296] 6. Notification of Results

[0297] The server notifies the user of the fact-check results received from the contributors and displays the results on the device. The results and data on the user's emotional state are also fed back as learning data for the AI ​​model, contributing to improving the accuracy of the system.

[0298] Hardware and Software

[0299] Key hardware and software used:

[0300] Smartphone: Camera and microphone for user interface and emotion recognition.

[0301] Server: Data management and fact-checking queue management.

[0302] Emotion Recognition API: Recognize user emotions in real time.

[0303] Fact Check API: Request fact checks and get results.

[0304] Notifications API: Notify users of fact-check results.

[0305] Specific examples

[0306] For example, suppose a user generates text such as "The highest mountain in the world is Mount Fuji" and has doubts while reading it. At this point, the emotion engine detects the user's emotions, and the system automatically requests a fact-check. The fact-check result, "The highest mountain in the world is Mount Everest," is provided and notified to the user's device.

[0307] Prompt Sentence Examples

[0308] text

[0309] If a user reads the article "Mount Fuji is the tallest mountain in the world" and has doubts, create a Python program to request a fact-check and display the results. Also, incorporate an engine to recognize the user's emotional state.

[0310] In this way, the present invention is a system that enables users to safely verify the authenticity of generated text and obtain highly reliable information.

[0311] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0312] Step 1:

[0313] A user accesses a generative AI system to generate text. The input is the content of the text the user wants to generate and a prompt, and the output is the generated text. The generative AI model generates text based on the prompt and displays it on the user's device.

[0314] Step 2:

[0315] The terminal displays the generated text and a button to request a fact check. This button is used when the user wants to verify the authenticity of the text. The input is the generated text and the state of the fact check button, and the output is the display of the button.

[0316] Step 3:

[0317] When a user presses the fact-check request button, the device sends the generated text to the server. At the same time, the emotion engine analyzes the user's emotional state and sends that information to the server. The input is the user's operation (button press) and emotional state data, and the output is data sent to the server.

[0318] Step 4:

[0319] The server adds the received generated text and emotional state information to a queue as a fact-checking request. At this time, the queue also contains metadata such as the date and time of the request's receipt and the emotional state. The input is the received data (generated text and emotional information), and the output is the fact-checking request added to the queue.

[0320] Step 5:

[0321] Fact-checking contributors access the server through a dedicated interface and retrieve fact-checking requests from the queue. The contributors check the contents of the generated text and conduct research to verify its authenticity. The input is the generated text retrieved from the queue, and the output is the fact-checking results.

[0322] Step 6:

[0323] The server receives the fact-check results returned by the contributors and stores them in a database. The saved fact-check results are displayed the next time the user accesses the server. The input is the fact-check results from the contributors, and the output is saving them in the database and notifying the user.

[0324] Step 7:

[0325] The server then feeds back the stored fact-check results and emotional state information as training data for the AI ​​model. This feedback improves the accuracy and reliability of the generative AI. The input is the fact-check results and emotional state information, and the output is an update to the AI ​​model.

[0326] Step 8:

[0327] The device receives a notification from the server and displays the generated fact-check result for the text, allowing the user to verify the authenticity of the information. The input is the fact-check result notification from the server, and the output is the result displayed on the user's device.

[0328] Through these steps, the system allows users to verify the authenticity of the generated text and realizes flexible fact-checking that takes into account emotional states.

[0329] 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.

[0330] 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.

[0331] 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.

[0332] [Second embodiment]

[0333] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0334] 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.

[0335] 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).

[0336] 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.

[0337] 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.

[0338] 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).

[0339] 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.

[0340] 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.

[0341] 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.

[0342] 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.

[0343] In the smart glasses 214, 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.

[0344] 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."

[0345] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0346] First, a user accesses the generative AI system using their own device and generates text. The generated text is immediately displayed on the user's device, with a button next to the text to request fact-checking.

[0347] When a user presses this button, the corresponding text is sent to the server. The server adds the received text to a queue as a fact-check request. After the fact-check request is added to the queue, a contributor retrieves the request from the queue and works to verify the authenticity of the text. The contributor then sends the verification result to the server, which receives and stores the result.

[0348] The server then notifies the user of the received fact-check results. Specifically, the results are sent to the user's device and displayed on the user interface. The fact-check results are also used as training data for the AI ​​model. The server feeds the results back into the AI ​​model to improve its accuracy and reliability.

[0349] Furthermore, contributors who provide excellent fact-checking results will be rewarded. This reward system is expected to increase contributors' motivation and lead to higher quality fact-checking.

[0350] As a specific example, consider the case where a user generates text such as "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to a queue. A collaborator retrieves this request from the queue and obtains the check result, "Mount Everest is the highest." The result is sent to the server, which notifies the user of the result and provides feedback to the AI ​​model.

[0351] Each process in this system is carried out as follows: The user checks the text generated by the generative AI on their own device and sends a request to the server via the fact-check button. The server adds the request to a queue, receives and stores the results from contributors, and notifies the user. Finally, feedback is provided to the AI ​​model and contributors are rewarded. This series of steps allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole.

[0352] The above is an embodiment of the present invention. This system allows users to verify the authenticity of text without any burden, and enables them to quickly obtain accurate information.

[0353] The processing flow will be explained below.

[0354] Step 1:

[0355] A user uses a generative AI system to request text generation. Based on this request, the server instructs the generative AI to generate text. The generated text is sent from the server to the user's device and displayed on the device's screen.

[0356] Step 2:

[0357] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the generated text.

[0358] Step 3:

[0359] The user clicks the "Outsource fact-checking" button. The device takes the generated text and sends a fact-checking request in JSON format to the server.

[0360] Step 4:

[0361] The server processes the received fact-check request by adding it to a queue and notifying the user that the request has been received.

[0362] Step 5:

[0363] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor uses their own resources and knowledge to verify the authenticity of the generated text.

[0364] Step 6:

[0365] The collaborators compile the verification results and send them to the server, which receives them and stores them in a database.

[0366] Step 7:

[0367] The server notifies the user of the saved fact-check results, which are then displayed on the user's device, allowing the user to view the results.

[0368] Step 8:

[0369] The server feeds the fact-checking results back into the AI ​​model, which then undergoes a learning process to improve the model's accuracy and reliability.

[0370] Step 9:

[0371] The server evaluates the quality of the fact-check results provided by contributors and rewards contributors who provide excellent results. The server manages the distribution of rewards and notifies contributors.

[0372] This allows users to efficiently verify the authenticity of generated text, improving the reliability of the entire system.

[0373] Example 1

[0374] 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."

[0375] Currently, many generative AI systems generate text for users, but the lack of a means to verify the authenticity of the generated text has led to the spread of misinformation. Furthermore, the fact-checking process is often done manually, which is time-consuming and the quality of the checks cannot be guaranteed. Furthermore, there is a lack of mechanisms to maintain the motivation of contributors and continuously conduct high-quality fact-checks. Therefore, there is a need for the development of a system that can quickly and accurately verify the authenticity of generated text.

[0376] 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.

[0377] In this invention, the server includes: means for displaying the generated text so that the user can verify the authenticity of the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a contributor to obtain the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the artificial intelligence; means for displaying the result on the user's terminal based on the generated text; and means for calculating and notifying a reward to the contributor. This allows the user to quickly and accurately verify the authenticity of the generated text, improving the reliability of the entire system.

[0378] A "user terminal" is a device used by a user to display the generated text and request fact-checking.

[0379] "Generative AI" is artificial intelligence that generates text in response to a prompt.

[0380] The "fact check button" is an operation unit that allows the user to request confirmation of the authenticity of the generated text.

[0381] The "server" is a device that receives fact-checking requests, adds them to a queue, notifies the user of the results, and provides feedback to the artificial intelligence.

[0382] A "queue" is a data structure that allows a server to manage fact-check requests sequentially.

[0383] A "collaborator" is an individual or organization that responds to a fact-check request and verifies the authenticity of the text.

[0384] "Fact-check results" are information about the authenticity of a text that has been confirmed by a contributor.

[0385] "Artificial intelligence training data" refers to fact-checking results data used to improve the accuracy and reliability of generative AI.

[0386] "Rewards" are compensation paid to contributors who provide excellent fact-checking results.

[0387] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0388] First, a user accesses the generative AI system using their own device and generates text. Specifically, they access the generative AI system's interface using a web browser. The user inputs a prompt sentence. For example, they might input "Mount Fuji is the highest mountain in the world." The generative AI model (e.g., OpenAI's GPT-3) generates text based on this prompt sentence. The generated text is immediately displayed on the user's device.

[0389] Next to the generated text is a button for requesting a fact check. When a user clicks this fact check button, the text is sent to the server as a fact check request. The server processes the received fact check request and adds it to a queue. This queue is a data structure for managing the processing order of requests.

[0390] The collaborator retrieves a request from the queue using a dedicated interface. The collaborator verifies the authenticity of the text content, for example, using the Internet or literature. For example, for the text "Mount Fuji is the highest mountain in the world," the collaborator obtains the result "Mount Everest is the highest." The verification result is entered into a dedicated input form and sent to the server.

[0391] The server stores the received fact-check results and notifies the user. The results are displayed on the user's device. For example, the result "Everest is the highest mountain" is displayed. The server also stores the fact-check results in a database for feedback to the generative AI model. This feedback improves the accuracy and reliability of the generative AI model.

[0392] In addition, contributors who provide excellent fact-checking results are rewarded. This reward can be paid via PayPal, for example. The server calculates the reward for contributors and notifies them.

[0393] As a concrete example of this system, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to the queue. A collaborator retrieves this request from the queue and obtains the check result "Mount Everest is the highest." The result is sent to the server, which notifies the user and provides feedback to the generative AI model.

[0394] This process allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole. Furthermore, by providing appropriate rewards to contributors, it is expected that high-quality fact-checking will be carried out sustainably. This concludes the description of the present invention.

[0395] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0396] Step 1:

[0397] A user accesses a generative AI system and inputs a prompt sentence. For example, they access the system's interface using a web browser. They input the prompt sentence, "Mount Fuji is the highest mountain in the world." Based on this input, a generative AI model (e.g., OpenAI's GPT-3) generates text. The generated text is displayed on the user's device screen. The input here is the prompt sentence, and the output is the generated text.

[0398] Step 2:

[0399] The user clicks on the fact-check button located next to the generated text. This click becomes the input, and a fact-check request is sent to the server along with the generated text. The server processes the received text as a fact-check request and adds it to a queue, which is a data structure. Here, the input is the user's click and the generated text, and the output is the server adding the fact-check request to the queue.

[0400] Step 3:

[0401] The server adds fact-check requests to a queue and monitors their status. Contributors retrieve requests from the queue using a dedicated interface. The retrieved text is sent to the contributor's device, and the contributor begins work to verify the text's content. The input here is the fact-check request added to the queue, and the output is the text received on the contributor's device.

[0402] Step 4:

[0403] A collaborator verifies the authenticity of the text content using the Internet or literature. For example, for a text that reads "Mount Fuji is the highest mountain in the world," the verification result is "Mount Everest is the highest mountain." The collaborator enters the verification result into a dedicated input form and sends it to the server. The input here is the text content and verification materials, and the output is the verification result.

[0404] Step 5:

[0405] The server stores the received confirmation result. The server notifies the user's terminal of this result. For example, the result "Everest is the highest mountain" is displayed on the user's terminal. The input here is the confirmation result, and the output is the notification to the user terminal.

[0406] Step 6:

[0407] The server feeds back the fact-check results to the generative AI model. The received results are stored in a database and used as training data for the generative AI model. The input here is the stored fact-check results, and the output is a reflection on the training data for the generative AI model.

[0408] Step 7:

[0409] The server calculates rewards for contributors. Contributors who provide good fact-check results are paid, for example, via PayPal. The payment of rewards includes a notification from the server to the contributor. The input here is the evaluation of the fact-check result, and the output is the payment of rewards to contributors.

[0410] (Application example 1)

[0411] 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."

[0412] The challenge is to quickly verify the reliability of text and product information generated by generative AI models, ensuring that users can safely use accurate information. It is also necessary to ensure the authenticity of product information provided in virtual stores, thereby improving users' shopping experiences.

[0413] 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.

[0414] In this invention, the server includes: means for a user to verify the authenticity of generated text; means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for an AI; means for providing product information to the user in a virtual store; means for displaying a button in the virtual store for requesting a fact-check on the generated product information; and means for requesting a fact-check through the product information displayed on the user's smart device. This allows users to quickly verify the authenticity of product information provided in the virtual store and enjoy shopping with peace of mind.

[0415] "User" refers to a person who uses a generative AI model to generate text or product information and verify the authenticity of that content.

[0416] "Generated text" refers to sentences or descriptions automatically created by a generative AI model.

[0417] "Server" refers to the computer system that manages the generated text and fact-checking requests, processes fact-checking results from contributors, and notifies users.

[0418] "Fact checking" refers to the process of verifying the authenticity of the content of generated text or product information.

[0419] A "collaborator" is someone who receives a fact-check request sent from the server, checks its contents, and sends the results to the server.

[0420] A "queue" refers to a data structure that stores received fact-check requests in order and provides them to contributors.

[0421] "AI training data" refers to data used to improve the accuracy and reliability of generative AI models.

[0422] "Virtual store" refers to a virtual store environment where products are sold and purchased over the Internet.

[0423] "Product information" refers to information such as product features, specifications, reviews, etc. presented in the virtual store.

[0424] "Smart devices" refers to portable electronic devices that can connect to the Internet, such as smartphones and smart glasses.

[0425] The present invention provides a system for outsourcing fact-checking to verify the authenticity of text and product information output by a generative AI model. The following describes in detail an embodiment of the present invention.

[0426] System Overview

[0427] The system consists of the following elements:

[0428] 1. User Device

[0429] 2. Cloud Server

[0430] 3. Collaborator terminal

[0431] 4. Generative AI Models

[0432] 5. Smart devices (smartphones, smart glasses)

[0433] System Operation

[0434] User device:

[0435] The user terminal provides a user interface for generating text and product information. The user inputs a prompt sentence into the generative AI model to generate the text and product information. The generated text and product information are displayed on the screen of the user terminal.

[0436] Cloud Server:

[0437] The cloud server plays a central role in managing fact-checking requests. When a user submits a fact-checking request, it is sent to the server and added to a queue. The server then provides the fact-checking requests to contributors and collects the results.

[0438] Collaborator's device:

[0439] The contributor terminal provides an interface for receiving fact-check requests. The contributor processes the requests and sends the results to the cloud server. Contributors are rewarded for providing good fact-check answers.

[0440] Generative AI models:

[0441] A generative AI model (e.g., GPT-4) generates text and product information based on a prompt entered by the user. This model runs on a cloud server and provides information in response to requests from user devices.

[0442] Smart Devices:

[0443] Users use smart devices (e.g., smartphones or smart glasses) to view the generated text and product information and submit fact-check requests.

[0444] Data processing and calculation

[0445] The cloud server processes data using the following software and hardware:

[0446] software:

[0447] Generative AI models (e.g., GPT-4)

[0448] Virtual store platforms (e.g., Shopify, WooCommerce)

[0449] Task management systems (e.g., Trello)

[0450] Hardware:

[0451] Cloud-based servers (e.g., Amazon Web Services) that provide computing power

[0452] Detailed procedure

[0453] 1. Information generation and display:

[0454] When a user accesses a product page in the virtual store, the generative AI model generates product information and displays it on the user's smart device.

[0455] 2. Fact-checking requests:

[0456] When a user presses the fact-check button, the generated text and product information are sent to the cloud server and added to the queue as a fact-check request.

[0457] 3. Collaborator verification:

[0458] The collaborator retrieves the request from the queue and checks the authenticity of the product information. The result is sent to the cloud server and notified to the user.

[0459] 4. Feedback of results:

[0460] The cloud server feeds the received results back into the AI ​​model to improve the model's accuracy.

[0461] Specific examples

[0462] For example, imagine a user wearing smart glasses accesses a virtual store and browses the product page for the latest smartphone. At this time, the generative AI model generates information such as "The latest smartphones are 5G compatible, have high-resolution cameras, and have long-lasting batteries." If the user wants to verify the authenticity of this information, they can click the fact-check button. The results of the check by the collaborators are then notified, and the user can receive reliable information such as "This smartphone is actually 5G compatible and has a long-lasting battery."

[0463] Prompt Sentence Examples

[0464] Generate user-specified product features and specifications.

[0465] Explain the effect of the fact-check button on product information.

[0466] This system allows users to quickly verify the authenticity of product information provided in virtual stores, allowing them to enjoy shopping with peace of mind.

[0467] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0468] Step 1:

[0469] The user device sends a request to the generative AI model.

[0470] Specifically, the user enters a prompt sentence and clicks the generate button. This sends the prompt sentence to the cloud server as an API request. The input data is the prompt sentence, and the output data is the generated text. The server invokes the generative AI model, generates text based on the entered prompt sentence, and sends the result to the user's device.

[0471] Step 2:

[0472] The user terminal displays the generated text.

[0473] The generated text received from the server is displayed on the user terminal. The user checks this text and determines its authenticity. The input data is the generated text received from the server, and the output data is the text displayed on the user interface.

[0474] Step 3:

[0475] The user terminal sends a fact-check request to the server.

[0476] When a user clicks the fact-check button, the generated text is sent to the cloud server as a fact-check request. The input data is the generated text and the fact-check request action, and the output data is the data to be added to the queue as a fact-check request. The server adds the request to a queue data structure.

[0477] Step 4:

[0478] The server provides fact-check requests to contributors.

[0479] The server retrieves fact-check requests from the queue and provides them to the collaborator terminal. The input data is the fact-check request in the queue, and the output data is the request sent to the collaborator terminal. The collaborator receives the request and starts checking it.

[0480] Step 5:

[0481] Contributors submit fact-check results to the server.

[0482] The contributor processes the request and sends the results to the cloud server. The input data is the fact-check request, and the output data is the verification result. The server stores the results received from the contributor and prepares them for further processing.

[0483] Step 6:

[0484] The server notifies the user terminal of the fact-check results.

[0485] The server sends the received fact-check results to the user's device. The input data is the confirmation result received from the collaborator, and the output data is a notification displayed on the user's device. The user can check the result.

[0486] Step 7:

[0487] The server feeds the fact-checking results back to the generative AI model.

[0488] The server feeds back the fact-check results as training data for the generative AI model. This feedback improves the accuracy and reliability of the AI ​​model. The input data is the fact-check results, and the output data is an updated AI model.

[0489] Step 8:

[0490] The user's smart device displays the product information.

[0491] The generated product information is displayed on the smart device, allowing for fact-checking requests. The input data is the generated product information, and the output data is the information displayed on the smart device. The user can click to submit the request.

[0492] 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.

[0493] The present invention is a system that outsources fact-checking to verify the authenticity of the content of text output by generative AI, and further combines it with an emotion engine that recognizes the emotional state of the user and controls the operation of the system based on that. Below, we will explain in detail the embodiments of the present invention.

[0494] First, the user accesses the generative AI system using their own device and generates text. This generated text is immediately displayed on the user's device. Next to the generated text, there is a button to request a fact check. This button is used if the user wants to verify the authenticity of the generated text.

[0495] Next, when the user presses this button, the corresponding text is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received text and emotional state information to a queue as a fact-check request and notifies the user of the receipt of the request.

[0496] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated text and sends the results to the server. The server receives the results and stores them in a database. The saved fact-check results are notified to the user and displayed on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[0497] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[0498] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[0499] As a specific example, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check result, and the server notifies the user of the result. At the same time, the result and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[0500] The above is an embodiment of the present invention. The introduction of an emotion engine enables flexible fact-checking that takes into account the emotional state of the user, further improving the usability and reliability of the system.

[0501] The processing flow will be explained below.

[0502] This invention is a system that outsources fact-checking to verify the authenticity of the text content output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that. The processing flow is explained below in specific steps.

[0503] Step 1:

[0504] A user accesses the generative AI system using a device and requests text generation. The server instructs the generative AI to generate text and sends the generated text to the user's device. The device then displays the generated text on its screen.

[0505] Step 2:

[0506] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the text.

[0507] Step 3:

[0508] The user clicks the "Outsource fact-checking" button. The device retrieves the generated text, and the emotion engine recognizes the user's current emotional state. This information is then sent to the server in JSON format as a fact-checking request.

[0509] Step 4:

[0510] The server adds the received fact-check request to a queue and notifies the user with a message confirming receipt of the request.

[0511] Step 5:

[0512] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor verifies the authenticity of the generated text and sends the result to the server.

[0513] Step 6:

[0514] The server receives fact-check results from contributors, stores them in a database, and notifies the user of the results and displays them on the user's device.

[0515] Step 7:

[0516] The server then feeds back the stored fact-check results and the user's emotional state information to the AI ​​model, which is used to improve the model's performance.

[0517] Step 8:

[0518] The emotion engine determines the need for fact-checking based on the user's emotional state. If the user expresses anxiety or doubt, the system automatically sends a fact-check request to the server, reducing the burden on the user.

[0519] Step 9:

[0520] The server evaluates the quality of the fact-check results provided by contributors, rewards contributors who provide excellent results, and manages the distribution of rewards and notifies contributors.

[0521] Through these steps, users can efficiently verify the authenticity of the generated text, improving the reliability and usability of the entire system. The introduction of an emotion engine enables flexible responses that take into account the user's psychological state, quickly ensuring the accuracy of information.

[0522] Example 2

[0523] 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."

[0524] There is a need for a method to quickly and accurately verify the authenticity of user-generated documents. Conventional systems have problems: fact-checking takes time and it is difficult to respond appropriately taking into account the user's emotional state. This can lead to users becoming anxious about the reliability of the information, which can lead to a decrease in the reliability of the system.

[0525] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for displaying a document created by a user, means for providing a button for requesting a fact-check for the document, means for transmitting the document to the server when a fact-check request is made by the user, means for adding the fact-check request received by the server to a queue, means for a collaborator to acquire the fact-check request from the queue, check the content, and transmit the result to the server, means for notifying the user of the fact-check result received by the server, means for feeding back the result and information on the user's emotional state as learning data for the AI, and an emotion engine that recognizes the user's emotional state and automatically requests a fact-check. This makes it possible to quickly and accurately verify the authenticity of a document while taking the user's emotional state into consideration.

[0526] A "user" is an entity that uses the system to generate documents and request fact-checking.

[0527] "Document" refers to text data created by generative AI.

[0528] "Generative AI" is an artificial intelligence model that generates documents based on prompts entered by the user.

[0529] "Fact checking" is a verification process to confirm the authenticity of the contents of a document.

[0530] A "button" is an interface element that a user clicks on the screen to request a specific action from the system.

[0531] A "server" is a computer system that processes and stores user requests and data.

[0532] A "queue" is a data structure for queuing received fact-check requests.

[0533] A "collaborator" is a person or system that receives a fact-check request, verifies its contents, and provides the results.

[0534] "Emotional state information" is data that represents the emotional state of the user.

[0535] "Training data" is a collection of past data and results used to improve the performance of an AI model.

[0536] An "emotion engine" is software or hardware that recognizes the user's emotional state and controls the system's behavior based on that.

[0537] "Feedback" refers to the re-input and re-processing of specific data or results for the purpose of improving the system.

[0538] "Notification" is a communication method by which a server provides information to a user.

[0539] A "prompt" refers to an instruction or question that is input to a generative AI.

[0540] This invention is a system that outsources fact-checking to verify the authenticity of the contents of documents output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that.

[0541] First, a user uses their own device to access a generative AI system (e.g., a generative AI model) and enters a prompt to generate a document. The generated document is immediately displayed on the user's device. Next to the generated document, there is a button for requesting fact-checking. This button is used when the user wants to verify the authenticity of the generated document.

[0542] Next, when the user presses the fact-check button, the corresponding document is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received document and emotional state information to a queue (e.g., AWS SQS) as a fact-check request and notifies the user of the receipt of the request.

[0543] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated document and sends the results to the server. The server receives the results and stores them in a database (e.g., MySQL). The server notifies the user of the stored fact-check results and displays them on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[0544] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[0545] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[0546] As a specific example, consider the case where a user generates a document stating "Mount Fuji is the tallest mountain in the world" and requests that it be fact-checked. This document and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check results, and the server notifies the user of the results. At the same time, the results and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[0547] The introduction of an emotion engine enables flexible fact-checking that takes into account the user's emotional state, further improving the usability and reliability of the system.

[0548] An example of a prompt sentence is "Generate the document 'Mount Fuji is the highest mountain in the world' and request a fact check."

[0549] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0550] Step 1:

[0551] A user accesses the generative AI system using their own device. The user inputs a prompt and requests text generation. The device sends the user's input (prompt) to the generative AI model and receives the generated text. For example, if a user inputs the prompt "Tell me about Mount Fuji," the device receives the text "Mount Fuji is the highest mountain in Japan" from the generative AI. This generated text is immediately displayed on the device.

[0552] Step 2:

[0553] The user presses a fact-check button displayed next to the generated text. The device sends the generated text and the user's emotional state (e.g., anxiety or doubt) to the server. The input here is the generated text and the user's emotional state information, which the device then sends to the server. For example, the device sends the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state information of anxiety.

[0554] Step 3:

[0555] The server adds the received text and emotional state information to a queue. As input, the server receives the generated text and emotional state information received from the device. The server adds this information to a queue (e.g., AWS SQS) and processes it as a fact-checking request. For example, the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state are added to the queue.

[0556] Step 4:

[0557] The server notifies the user that it has received the fact-checking request. The server confirms that the fact-checking request has been added to the queue and notifies the user that it has been received. The input is the queue information added to the server, and the output is a notification to the user. For example, a notification saying "your fact-checking request has been accepted" is sent to the user.

[0558] Step 5:

[0559] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. As input, the contributor retrieves a fact-check request from the queue, and as output, the retrieved text information is obtained. For example, a contributor retrieves the text "Mount Fuji is the highest mountain in Japan."

[0560] Step 6:

[0561] The collaborator verifies the truth of the generated text content and sends the result to the server. As input, the collaborator receives the text information to be verified and investigates its content. As output, the investigation result is sent to the server. For example, the collaborator verifies the result that "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" and sends it to the server.

[0562] Step 7:

[0563] The server receives the fact-check results and stores them in a database. As input, it receives the fact-check results sent by the contributors, and as output, it stores them in a database. For example, the result "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" is stored in the database.

[0564] Step 8:

[0565] The server notifies the user of the saved fact-check results and displays the results on the user's device. As input, the server retrieves the saved results from the database and notifies the user as output. The user can view the fact-check results on their device. For example, a notification is displayed to the user saying, "Fact-check result: Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world."

[0566] Step 9:

[0567] The server feeds back the fact-check results and the user's emotional state information as learning data for the AI ​​model. The fact-check results and emotional state information are used as input, and data to improve the accuracy of the AI ​​model is obtained as output. For example, the text "Mount Fuji is the highest mountain in Japan," the emotional state "anxiety," and the fact-check results are fed back to the AI ​​model.

[0568] Step 10:

[0569] The emotion engine recognizes the user's emotional state and automatically requests fact-checking when necessary. The user's emotional state information is used as input, and an automatic fact-checking request is generated as output. For example, if the user expresses anxiety, the system will automatically request a fact-check.

[0570] By using the above steps, the system of the present invention can quickly and accurately verify the authenticity of a document while taking into account the emotional state of the user.

[0571] (Application example 2)

[0572] 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."

[0573] In recent years, text generation using generative AI has become widespread, and the importance of fact-checking to verify the authenticity of generated text is increasing. However, the process for users to verify the reliability of generated text is time-consuming and often leaves users feeling anxious or suspicious. Furthermore, a system that takes into account the user's emotional state is needed to efficiently verify the authenticity of generated text. Furthermore, a reward system for contributors is also necessary to improve the quality of fact-checking results. The present invention aims to provide a system that solves these problems.

[0574] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a user requests a fact-check; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, check the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the AI; means for recognizing the user's emotional state using an emotion engine and controlling the system operation based on that information; means for automatically requesting a fact-check according to the emotional state; and means for displaying the fact-check result based on the emotional state on the user's terminal. This allows the user to quickly and efficiently verify the authenticity of the generated text, thereby realizing flexible fact-checking that takes the user's emotional state into consideration.

[0575] "User" means an individual or organization that uses the generative AI system.

[0576] "Generated text" refers to textual information generated by a generation AI.

[0577] "Fact checking" is the act of verifying the truth of the content of the generated text.

[0578] The "emotion engine" is an engine that recognizes the user's emotional state and reflects that information in the system's operation.

[0579] "Server" means the computer system that manages and processes data and fact-checking requests submitted by users.

[0580] A "collaborator" is an individual or organization whose role is to receive a fact-check request, verify its contents, and return the results.

[0581] A "queue" is a data structure for lining up received fact-check requests in order and waiting.

[0582] "Notification" means the act or means of communicating fact-check results or other information to a user.

[0583] "Training data" is data used to train and improve AI models.

[0584] The "reward system" is a mechanism that rewards contributors for providing excellent fact-checking results.

[0585] "Generative AI" is artificial intelligence that allows users to generate text.

[0586] A "terminal" is a device through which a user accesses the generative AI system.

[0587] The present invention provides a system for accessing a generation AI system on a user's terminal and verifying the authenticity of generated text. Specific embodiments for carrying out the present invention will be described below.

[0588] Overall system configuration

[0589] The system consists of the following main components:

[0590] 1. User devices: smartphones, tablets, computers, etc.

[0591] 2. Server: Manages the queue of fact-checking requests and stores the results.

[0592] 3. Generative AI: An AI engine that allows users to generate text.

[0593] 4. Emotion engine: Recognizes the user's emotional state and controls the system behavior.

[0594] 5. Fact-checking contributors: These contributors are responsible for verifying the authenticity of the generated text.

[0595] Program Overview

[0596] 1. Text Generation and Display

[0597] Users can use generative AI through their device to generate specific text, which is then instantly displayed on the device.

[0598] 2. Recognizing emotional states

[0599] The emotion engine uses the device's camera, microphone, and sensors to recognize the user's emotional state in real time, for example, using facial recognition and voice analysis technologies.

[0600] 3. Fact-checking requests

[0601] A button to request a fact check will be displayed next to the generated text. If the user has any concerns or doubts, they can press this button to send a fact check request to the server.

[0602] 4. Queue Management

[0603] The server adds the received fact-check request to a queue, and at the same time, the emotional state data sent from the emotion engine is also recorded on the server.

[0604] 5. Fact Check

[0605] Contributors access the server using a dedicated interface to retrieve fact-check requests added to the queue, check their contents, and send the results to the server.

[0606] 6. Notification of Results

[0607] The server notifies the user of the fact-check results received from the contributors and displays the results on the device. The results and data on the user's emotional state are also fed back as learning data for the AI ​​model, contributing to improving the accuracy of the system.

[0608] Hardware and Software

[0609] Key hardware and software used:

[0610] Smartphone: Camera and microphone for user interface and emotion recognition.

[0611] Server: Data management and fact-checking queue management.

[0612] Emotion Recognition API: Recognize user emotions in real time.

[0613] Fact Check API: Request fact checks and get results.

[0614] Notifications API: Notify users of fact-check results.

[0615] Specific examples

[0616] For example, suppose a user generates text such as "The highest mountain in the world is Mount Fuji" and has doubts while reading it. At this point, the emotion engine detects the user's emotions, and the system automatically requests a fact-check. The fact-check result, "The highest mountain in the world is Mount Everest," is provided and notified to the user's device.

[0617] Prompt Sentence Examples

[0618] text

[0619] If a user reads the article "Mount Fuji is the tallest mountain in the world" and has doubts, create a Python program to request a fact-check and display the results. Also, incorporate an engine to recognize the user's emotional state.

[0620] In this way, the present invention is a system that enables users to safely verify the authenticity of generated text and obtain highly reliable information.

[0621] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0622] Step 1:

[0623] A user accesses a generative AI system to generate text. The input is the content of the text the user wants to generate and a prompt, and the output is the generated text. The generative AI model generates text based on the prompt and displays it on the user's device.

[0624] Step 2:

[0625] The terminal displays the generated text and a button to request a fact check. This button is used when the user wants to verify the authenticity of the text. The input is the generated text and the state of the fact check button, and the output is the display of the button.

[0626] Step 3:

[0627] When a user presses the fact-check request button, the device sends the generated text to the server. At the same time, the emotion engine analyzes the user's emotional state and sends that information to the server. The input is the user's operation (button press) and emotional state data, and the output is data sent to the server.

[0628] Step 4:

[0629] The server adds the received generated text and emotional state information to a queue as a fact-checking request. At this time, the queue also contains metadata such as the date and time of the request's receipt and the emotional state. The input is the received data (generated text and emotional information), and the output is the fact-checking request added to the queue.

[0630] Step 5:

[0631] Fact-checking contributors access the server through a dedicated interface and retrieve fact-checking requests from the queue. The contributors check the contents of the generated text and conduct research to verify its authenticity. The input is the generated text retrieved from the queue, and the output is the fact-checking results.

[0632] Step 6:

[0633] The server receives the fact-check results returned by the contributors and stores them in a database. The saved fact-check results are displayed the next time the user accesses the server. The input is the fact-check results from the contributors, and the output is saving them in the database and notifying the user.

[0634] Step 7:

[0635] The server then feeds back the stored fact-check results and emotional state information as training data for the AI ​​model. This feedback improves the accuracy and reliability of the generative AI. The input is the fact-check results and emotional state information, and the output is an update to the AI ​​model.

[0636] Step 8:

[0637] The device receives a notification from the server and displays the generated fact-check result for the text, allowing the user to verify the authenticity of the information. The input is the fact-check result notification from the server, and the output is the result displayed on the user's device.

[0638] Through these steps, the system allows users to verify the authenticity of the generated text and realizes flexible fact-checking that takes into account emotional states.

[0639] 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.

[0640] 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.

[0641] 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.

[0642] [Third embodiment]

[0643] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0644] 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.

[0645] 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).

[0646] 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.

[0647] 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.

[0648] 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).

[0649] 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.

[0650] 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.

[0651] 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.

[0652] 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.

[0653] 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.

[0654] 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."

[0655] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0656] First, a user accesses the generative AI system using their own device and generates text. The generated text is immediately displayed on the user's device, with a button next to the text to request fact-checking.

[0657] When a user presses this button, the corresponding text is sent to the server. The server adds the received text to a queue as a fact-check request. After the fact-check request is added to the queue, a contributor retrieves the request from the queue and works to verify the authenticity of the text. The contributor then sends the verification result to the server, which receives and stores the result.

[0658] The server then notifies the user of the received fact-check results. Specifically, the results are sent to the user's device and displayed on the user interface. The fact-check results are also used as training data for the AI ​​model. The server feeds the results back into the AI ​​model to improve its accuracy and reliability.

[0659] Furthermore, contributors who provide excellent fact-checking results will be rewarded. This reward system is expected to increase contributors' motivation and lead to higher quality fact-checking.

[0660] As a specific example, consider the case where a user generates text such as "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to a queue. A collaborator retrieves this request from the queue and obtains the check result, "Mount Everest is the highest." The result is sent to the server, which notifies the user of the result and provides feedback to the AI ​​model.

[0661] Each process in this system is carried out as follows: The user checks the text generated by the generative AI on their own device and sends a request to the server via the fact-check button. The server adds the request to a queue, receives and stores the results from contributors, and notifies the user. Finally, feedback is provided to the AI ​​model and contributors are rewarded. This series of steps allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole.

[0662] The above is an embodiment of the present invention. This system allows users to verify the authenticity of text without any burden, and enables them to quickly obtain accurate information.

[0663] The processing flow will be explained below.

[0664] Step 1:

[0665] A user uses a generative AI system to request text generation. Based on this request, the server instructs the generative AI to generate text. The generated text is sent from the server to the user's device and displayed on the device's screen.

[0666] Step 2:

[0667] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the generated text.

[0668] Step 3:

[0669] The user clicks the "Outsource fact-checking" button. The device takes the generated text and sends a fact-checking request in JSON format to the server.

[0670] Step 4:

[0671] The server processes the received fact-check request by adding it to a queue and notifying the user that the request has been received.

[0672] Step 5:

[0673] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor uses their own resources and knowledge to verify the authenticity of the generated text.

[0674] Step 6:

[0675] The collaborators compile the verification results and send them to the server, which receives them and stores them in a database.

[0676] Step 7:

[0677] The server notifies the user of the saved fact-check results, which are then displayed on the user's device, allowing the user to view the results.

[0678] Step 8:

[0679] The server feeds the fact-checking results back into the AI ​​model, which then undergoes a learning process to improve the model's accuracy and reliability.

[0680] Step 9:

[0681] The server evaluates the quality of the fact-check results provided by contributors and rewards contributors who provide excellent results. The server manages the distribution of rewards and notifies contributors.

[0682] This allows users to efficiently verify the authenticity of generated text, improving the reliability of the entire system.

[0683] Example 1

[0684] 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."

[0685] Currently, many generative AI systems generate text for users, but the lack of a means to verify the authenticity of the generated text has led to the spread of misinformation. Furthermore, the fact-checking process is often done manually, which is time-consuming and the quality of the checks cannot be guaranteed. Furthermore, there is a lack of mechanisms to maintain the motivation of contributors and continuously conduct high-quality fact-checks. Therefore, there is a need for the development of a system that can quickly and accurately verify the authenticity of generated text.

[0686] 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.

[0687] In this invention, the server includes: means for displaying the generated text so that the user can verify the authenticity of the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a contributor to obtain the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the artificial intelligence; means for displaying the result on the user's terminal based on the generated text; and means for calculating and notifying a reward to the contributor. This allows the user to quickly and accurately verify the authenticity of the generated text, improving the reliability of the entire system.

[0688] A "user terminal" is a device used by a user to display the generated text and request fact-checking.

[0689] "Generative AI" is artificial intelligence that generates text in response to a prompt.

[0690] The "fact check button" is an operation unit that allows the user to request confirmation of the authenticity of the generated text.

[0691] The "server" is a device that receives fact-checking requests, adds them to a queue, notifies the user of the results, and provides feedback to the artificial intelligence.

[0692] A "queue" is a data structure that allows a server to manage fact-check requests sequentially.

[0693] A "collaborator" is an individual or organization that responds to a fact-check request and verifies the authenticity of the text.

[0694] "Fact-check results" are information about the authenticity of a text that has been confirmed by a contributor.

[0695] "Artificial intelligence training data" refers to fact-checking results data used to improve the accuracy and reliability of generative AI.

[0696] "Rewards" are compensation paid to contributors who provide excellent fact-checking results.

[0697] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0698] First, a user accesses the generative AI system using their own device and generates text. Specifically, they access the generative AI system's interface using a web browser. The user inputs a prompt sentence. For example, they might input "Mount Fuji is the highest mountain in the world." The generative AI model (e.g., OpenAI's GPT-3) generates text based on this prompt sentence. The generated text is immediately displayed on the user's device.

[0699] Next to the generated text is a button for requesting a fact check. When a user clicks this fact check button, the text is sent to the server as a fact check request. The server processes the received fact check request and adds it to a queue. This queue is a data structure for managing the processing order of requests.

[0700] The collaborator retrieves a request from the queue using a dedicated interface. The collaborator verifies the authenticity of the text content, for example, using the Internet or literature. For example, for the text "Mount Fuji is the highest mountain in the world," the collaborator obtains the result "Mount Everest is the highest." The verification result is entered into a dedicated input form and sent to the server.

[0701] The server stores the received fact-check results and notifies the user. The results are displayed on the user's device. For example, the result "Everest is the highest mountain" is displayed. The server also stores the fact-check results in a database for feedback to the generative AI model. This feedback improves the accuracy and reliability of the generative AI model.

[0702] In addition, contributors who provide excellent fact-checking results are rewarded. This reward can be paid via PayPal, for example. The server calculates the reward for contributors and notifies them.

[0703] As a concrete example of this system, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to the queue. A collaborator retrieves this request from the queue and obtains the check result "Mount Everest is the highest." The result is sent to the server, which notifies the user and provides feedback to the generative AI model.

[0704] This process allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole. Furthermore, by providing appropriate rewards to contributors, it is expected that high-quality fact-checking will be carried out sustainably. This concludes the description of the present invention.

[0705] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0706] Step 1:

[0707] A user accesses a generative AI system and inputs a prompt sentence. For example, they access the system's interface using a web browser. They input the prompt sentence, "Mount Fuji is the highest mountain in the world." Based on this input, a generative AI model (e.g., OpenAI's GPT-3) generates text. The generated text is displayed on the user's device screen. The input here is the prompt sentence, and the output is the generated text.

[0708] Step 2:

[0709] The user clicks on the fact-check button located next to the generated text. This click becomes the input, and a fact-check request is sent to the server along with the generated text. The server processes the received text as a fact-check request and adds it to a queue, which is a data structure. Here, the input is the user's click and the generated text, and the output is the server adding the fact-check request to the queue.

[0710] Step 3:

[0711] The server adds fact-check requests to a queue and monitors their status. Contributors retrieve requests from the queue using a dedicated interface. The retrieved text is sent to the contributor's device, and the contributor begins work to verify the text's content. The input here is the fact-check request added to the queue, and the output is the text received on the contributor's device.

[0712] Step 4:

[0713] A collaborator verifies the authenticity of the text content using the Internet or literature. For example, for a text that reads "Mount Fuji is the highest mountain in the world," the verification result is "Mount Everest is the highest mountain." The collaborator enters the verification result into a dedicated input form and sends it to the server. The input here is the text content and verification materials, and the output is the verification result.

[0714] Step 5:

[0715] The server stores the received confirmation result. The server notifies the user's terminal of this result. For example, the result "Everest is the highest mountain" is displayed on the user's terminal. The input here is the confirmation result, and the output is the notification to the user terminal.

[0716] Step 6:

[0717] The server feeds back the fact-check results to the generative AI model. The received results are stored in a database and used as training data for the generative AI model. The input here is the stored fact-check results, and the output is a reflection on the training data for the generative AI model.

[0718] Step 7:

[0719] The server calculates rewards for contributors. Contributors who provide good fact-check results are paid, for example, via PayPal. The payment of rewards includes a notification from the server to the contributor. The input here is the evaluation of the fact-check result, and the output is the payment of rewards to contributors.

[0720] (Application example 1)

[0721] 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."

[0722] The challenge is to quickly verify the reliability of text and product information generated by generative AI models, ensuring that users can safely use accurate information. It is also necessary to ensure the authenticity of product information provided in virtual stores, thereby improving users' shopping experiences.

[0723] 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.

[0724] In this invention, the server includes: means for a user to verify the authenticity of generated text; means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for an AI; means for providing product information to the user in a virtual store; means for displaying a button in the virtual store for requesting a fact-check on the generated product information; and means for requesting a fact-check through the product information displayed on the user's smart device. This allows users to quickly verify the authenticity of product information provided in the virtual store and enjoy shopping with peace of mind.

[0725] "User" refers to a person who uses a generative AI model to generate text or product information and verify the authenticity of that content.

[0726] "Generated text" refers to sentences or descriptions automatically created by a generative AI model.

[0727] "Server" refers to the computer system that manages the generated text and fact-checking requests, processes fact-checking results from contributors, and notifies users.

[0728] "Fact checking" refers to the process of verifying the authenticity of the content of generated text or product information.

[0729] A "collaborator" is someone who receives a fact-check request sent from the server, checks its contents, and sends the results to the server.

[0730] A "queue" refers to a data structure that stores received fact-check requests in order and provides them to contributors.

[0731] "AI training data" refers to data used to improve the accuracy and reliability of generative AI models.

[0732] "Virtual store" refers to a virtual store environment where products are sold and purchased over the Internet.

[0733] "Product information" refers to information such as product features, specifications, reviews, etc. presented in the virtual store.

[0734] "Smart devices" refers to portable electronic devices that can connect to the Internet, such as smartphones and smart glasses.

[0735] The present invention provides a system for outsourcing fact-checking to verify the authenticity of text and product information output by a generative AI model. The following describes in detail an embodiment of the present invention.

[0736] System Overview

[0737] The system consists of the following elements:

[0738] 1. User Device

[0739] 2. Cloud Server

[0740] 3. Collaborator terminal

[0741] 4. Generative AI Models

[0742] 5. Smart devices (smartphones, smart glasses)

[0743] System Operation

[0744] User device:

[0745] The user terminal provides a user interface for generating text and product information. The user inputs a prompt sentence into the generative AI model to generate the text and product information. The generated text and product information are displayed on the screen of the user terminal.

[0746] Cloud Server:

[0747] The cloud server plays a central role in managing fact-checking requests. When a user submits a fact-checking request, it is sent to the server and added to a queue. The server then provides the fact-checking requests to contributors and collects the results.

[0748] Collaborator's device:

[0749] The contributor terminal provides an interface for receiving fact-check requests. The contributor processes the requests and sends the results to the cloud server. Contributors are rewarded for providing good fact-check answers.

[0750] Generative AI models:

[0751] A generative AI model (e.g., GPT-4) generates text and product information based on a prompt entered by the user. This model runs on a cloud server and provides information in response to requests from user devices.

[0752] Smart Devices:

[0753] Users use smart devices (e.g., smartphones or smart glasses) to view the generated text and product information and submit fact-check requests.

[0754] Data processing and calculation

[0755] The cloud server processes data using the following software and hardware:

[0756] software:

[0757] Generative AI models (e.g., GPT-4)

[0758] Virtual store platforms (e.g., Shopify, WooCommerce)

[0759] Task management systems (e.g., Trello)

[0760] Hardware:

[0761] Cloud-based servers (e.g., Amazon Web Services) that provide computing power

[0762] Detailed procedure

[0763] 1. Information generation and display:

[0764] When a user accesses a product page in the virtual store, the generative AI model generates product information and displays it on the user's smart device.

[0765] 2. Fact-checking requests:

[0766] When a user presses the fact-check button, the generated text and product information are sent to the cloud server and added to the queue as a fact-check request.

[0767] 3. Collaborator verification:

[0768] The collaborator retrieves the request from the queue and checks the authenticity of the product information. The result is sent to the cloud server and notified to the user.

[0769] 4. Feedback of results:

[0770] The cloud server feeds the received results back into the AI ​​model to improve the model's accuracy.

[0771] Specific examples

[0772] For example, imagine a user wearing smart glasses accesses a virtual store and browses the product page for the latest smartphone. At this time, the generative AI model generates information such as "The latest smartphones are 5G compatible, have high-resolution cameras, and have long-lasting batteries." If the user wants to verify the authenticity of this information, they can click the fact-check button. The results of the check by the collaborators are then notified, and the user can receive reliable information such as "This smartphone is actually 5G compatible and has a long-lasting battery."

[0773] Prompt Sentence Examples

[0774] Generate user-specified product features and specifications.

[0775] Explain the effect of the fact-check button on product information.

[0776] This system allows users to quickly verify the authenticity of product information provided in virtual stores, allowing them to enjoy shopping with peace of mind.

[0777] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0778] Step 1:

[0779] The user device sends a request to the generative AI model.

[0780] Specifically, the user enters a prompt sentence and clicks the generate button. This sends the prompt sentence to the cloud server as an API request. The input data is the prompt sentence, and the output data is the generated text. The server invokes the generative AI model, generates text based on the entered prompt sentence, and sends the result to the user's device.

[0781] Step 2:

[0782] The user terminal displays the generated text.

[0783] The generated text received from the server is displayed on the user terminal. The user checks this text and determines its authenticity. The input data is the generated text received from the server, and the output data is the text displayed on the user interface.

[0784] Step 3:

[0785] The user terminal sends a fact-check request to the server.

[0786] When a user clicks the fact-check button, the generated text is sent to the cloud server as a fact-check request. The input data is the generated text and the fact-check request action, and the output data is the data to be added to the queue as a fact-check request. The server adds the request to a queue data structure.

[0787] Step 4:

[0788] The server provides fact-check requests to contributors.

[0789] The server retrieves fact-check requests from the queue and provides them to the collaborator terminal. The input data is the fact-check request in the queue, and the output data is the request sent to the collaborator terminal. The collaborator receives the request and starts checking it.

[0790] Step 5:

[0791] Contributors submit fact-check results to the server.

[0792] The contributor processes the request and sends the results to the cloud server. The input data is the fact-check request, and the output data is the verification result. The server stores the results received from the contributor and prepares them for further processing.

[0793] Step 6:

[0794] The server notifies the user terminal of the fact-check results.

[0795] The server sends the received fact-check results to the user's device. The input data is the confirmation result received from the collaborator, and the output data is a notification displayed on the user's device. The user can check the result.

[0796] Step 7:

[0797] The server feeds the fact-checking results back to the generative AI model.

[0798] The server feeds back the fact-check results as training data for the generative AI model. This feedback improves the accuracy and reliability of the AI ​​model. The input data is the fact-check results, and the output data is an updated AI model.

[0799] Step 8:

[0800] The user's smart device displays the product information.

[0801] The generated product information is displayed on the smart device, allowing for fact-checking requests. The input data is the generated product information, and the output data is the information displayed on the smart device. The user can click to submit the request.

[0802] 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.

[0803] The present invention is a system that outsources fact-checking to verify the authenticity of the content of text output by generative AI, and further combines it with an emotion engine that recognizes the emotional state of the user and controls the operation of the system based on that. Below, we will explain in detail the embodiments of the present invention.

[0804] First, the user accesses the generative AI system using their own device and generates text. This generated text is immediately displayed on the user's device. Next to the generated text, there is a button to request a fact check. This button is used if the user wants to verify the authenticity of the generated text.

[0805] Next, when the user presses this button, the corresponding text is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received text and emotional state information to a queue as a fact-check request and notifies the user of the receipt of the request.

[0806] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated text and sends the results to the server. The server receives the results and stores them in a database. The saved fact-check results are notified to the user and displayed on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[0807] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[0808] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[0809] As a specific example, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check result, and the server notifies the user of the result. At the same time, the result and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[0810] The above is an embodiment of the present invention. The introduction of an emotion engine enables flexible fact-checking that takes into account the emotional state of the user, further improving the usability and reliability of the system.

[0811] The processing flow will be explained below.

[0812] This invention is a system that outsources fact-checking to verify the authenticity of the text content output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that. The processing flow is explained below in specific steps.

[0813] Step 1:

[0814] A user accesses the generative AI system using a device and requests text generation. The server instructs the generative AI to generate text and sends the generated text to the user's device. The device then displays the generated text on its screen.

[0815] Step 2:

[0816] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the text.

[0817] Step 3:

[0818] The user clicks the "Outsource fact-checking" button. The device retrieves the generated text, and the emotion engine recognizes the user's current emotional state. This information is then sent to the server in JSON format as a fact-checking request.

[0819] Step 4:

[0820] The server adds the received fact-check request to a queue and notifies the user with a message confirming receipt of the request.

[0821] Step 5:

[0822] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor verifies the authenticity of the generated text and sends the result to the server.

[0823] Step 6:

[0824] The server receives fact-check results from contributors, stores them in a database, and notifies the user of the results and displays them on the user's device.

[0825] Step 7:

[0826] The server then feeds back the stored fact-check results and the user's emotional state information to the AI ​​model, which is used to improve the model's performance.

[0827] Step 8:

[0828] The emotion engine determines the need for fact-checking based on the user's emotional state. If the user expresses anxiety or doubt, the system automatically sends a fact-check request to the server, reducing the burden on the user.

[0829] Step 9:

[0830] The server evaluates the quality of the fact-check results provided by contributors, rewards contributors who provide excellent results, and manages the distribution of rewards and notifies contributors.

[0831] Through these steps, users can efficiently verify the authenticity of the generated text, improving the reliability and usability of the entire system. The introduction of an emotion engine enables flexible responses that take into account the user's psychological state, quickly ensuring the accuracy of information.

[0832] Example 2

[0833] 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."

[0834] There is a need for a method to quickly and accurately verify the authenticity of user-generated documents. Conventional systems have problems: fact-checking takes time and it is difficult to respond appropriately taking into account the user's emotional state. This can lead to users becoming anxious about the reliability of the information, which can lead to a decrease in the reliability of the system.

[0835] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for displaying a document created by a user, means for providing a button for requesting a fact-check for the document, means for transmitting the document to the server when a fact-check request is made by the user, means for adding the fact-check request received by the server to a queue, means for a collaborator to acquire the fact-check request from the queue, check the content, and transmit the result to the server, means for notifying the user of the fact-check result received by the server, means for feeding back the result and information on the user's emotional state as learning data for the AI, and an emotion engine that recognizes the user's emotional state and automatically requests a fact-check. This makes it possible to quickly and accurately verify the authenticity of a document while taking the user's emotional state into consideration.

[0836] A "user" is an entity that uses the system to generate documents and request fact-checking.

[0837] "Document" refers to text data created by generative AI.

[0838] "Generative AI" is an artificial intelligence model that generates documents based on prompts entered by the user.

[0839] "Fact checking" is a verification process to confirm the authenticity of the contents of a document.

[0840] A "button" is an interface element that a user clicks on the screen to request a specific action from the system.

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

[0842] A "queue" is a data structure for queuing received fact-check requests.

[0843] A "collaborator" is a person or system that receives a fact-check request, verifies its contents, and provides the results.

[0844] "Emotional state information" is data that represents the emotional state of the user.

[0845] "Training data" is a collection of past data and results used to improve the performance of an AI model.

[0846] An "emotion engine" is software or hardware that recognizes the user's emotional state and controls the system's behavior based on that.

[0847] "Feedback" refers to the re-input and re-processing of specific data or results for the purpose of improving the system.

[0848] "Notification" is a communication method by which a server provides information to a user.

[0849] A "prompt" refers to an instruction or question that is input to a generative AI.

[0850] This invention is a system that outsources fact-checking to verify the authenticity of the contents of documents output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that.

[0851] First, a user uses their own device to access a generative AI system (e.g., a generative AI model) and enters a prompt to generate a document. The generated document is immediately displayed on the user's device. Next to the generated document, there is a button for requesting fact-checking. This button is used when the user wants to verify the authenticity of the generated document.

[0852] Next, when the user presses the fact-check button, the corresponding document is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received document and emotional state information to a queue (e.g., AWS SQS) as a fact-check request and notifies the user of the receipt of the request.

[0853] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated document and sends the results to the server. The server receives the results and stores them in a database (e.g., MySQL). The server notifies the user of the stored fact-check results and displays them on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[0854] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[0855] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[0856] As a specific example, consider the case where a user generates a document stating "Mount Fuji is the tallest mountain in the world" and requests that it be fact-checked. This document and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check results, and the server notifies the user of the results. At the same time, the results and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[0857] The introduction of an emotion engine enables flexible fact-checking that takes into account the user's emotional state, further improving the usability and reliability of the system.

[0858] An example of a prompt sentence is "Generate the document 'Mount Fuji is the highest mountain in the world' and request a fact check."

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

[0860] Step 1:

[0861] A user accesses the generative AI system using their own device. The user inputs a prompt and requests text generation. The device sends the user's input (prompt) to the generative AI model and receives the generated text. For example, if a user inputs the prompt "Tell me about Mount Fuji," the device receives the text "Mount Fuji is the highest mountain in Japan" from the generative AI. This generated text is immediately displayed on the device.

[0862] Step 2:

[0863] The user presses a fact-check button displayed next to the generated text. The device sends the generated text and the user's emotional state (e.g., anxiety or doubt) to the server. The input here is the generated text and the user's emotional state information, which the device then sends to the server. For example, the device sends the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state information of anxiety.

[0864] Step 3:

[0865] The server adds the received text and emotional state information to a queue. As input, the server receives the generated text and emotional state information received from the device. The server adds this information to a queue (e.g., AWS SQS) and processes it as a fact-checking request. For example, the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state are added to the queue.

[0866] Step 4:

[0867] The server notifies the user that it has received the fact-checking request. The server confirms that the fact-checking request has been added to the queue and notifies the user that it has been received. The input is the queue information added to the server, and the output is a notification to the user. For example, a notification saying "your fact-checking request has been accepted" is sent to the user.

[0868] Step 5:

[0869] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. As input, the contributor retrieves a fact-check request from the queue, and as output, the retrieved text information is obtained. For example, a contributor retrieves the text "Mount Fuji is the highest mountain in Japan."

[0870] Step 6:

[0871] The collaborator verifies the truth of the generated text content and sends the result to the server. As input, the collaborator receives the text information to be verified and investigates its content. As output, the investigation result is sent to the server. For example, the collaborator verifies the result that "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" and sends it to the server.

[0872] Step 7:

[0873] The server receives the fact-check results and stores them in a database. As input, it receives the fact-check results sent by the contributors, and as output, it stores them in a database. For example, the result "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" is stored in the database.

[0874] Step 8:

[0875] The server notifies the user of the saved fact-check results and displays the results on the user's device. As input, the server retrieves the saved results from the database and notifies the user as output. The user can view the fact-check results on their device. For example, a notification is displayed to the user saying, "Fact-check result: Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world."

[0876] Step 9:

[0877] The server feeds back the fact-check results and the user's emotional state information as learning data for the AI ​​model. The fact-check results and emotional state information are used as input, and data to improve the accuracy of the AI ​​model is obtained as output. For example, the text "Mount Fuji is the highest mountain in Japan," the emotional state "anxiety," and the fact-check results are fed back to the AI ​​model.

[0878] Step 10:

[0879] The emotion engine recognizes the user's emotional state and automatically requests fact-checking when necessary. The user's emotional state information is used as input, and an automatic fact-checking request is generated as output. For example, if the user expresses anxiety, the system will automatically request a fact-check.

[0880] By using the above steps, the system of the present invention can quickly and accurately verify the authenticity of a document while taking into account the emotional state of the user.

[0881] (Application example 2)

[0882] 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."

[0883] In recent years, text generation using generative AI has become widespread, and the importance of fact-checking to verify the authenticity of generated text is increasing. However, the process for users to verify the reliability of generated text is time-consuming and often leaves users feeling anxious or suspicious. Furthermore, a system that takes into account the user's emotional state is needed to efficiently verify the authenticity of generated text. Furthermore, a reward system for contributors is also necessary to improve the quality of fact-checking results. The present invention aims to provide a system that solves these problems.

[0884] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a user requests a fact-check; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, check the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the AI; means for recognizing the user's emotional state using an emotion engine and controlling the system operation based on that information; means for automatically requesting a fact-check according to the emotional state; and means for displaying the fact-check result based on the emotional state on the user's terminal. This allows the user to quickly and efficiently verify the authenticity of the generated text, thereby realizing flexible fact-checking that takes the user's emotional state into consideration.

[0885] "User" means an individual or organization that uses the generative AI system.

[0886] "Generated text" refers to textual information generated by a generation AI.

[0887] "Fact checking" is the act of verifying the truth of the content of the generated text.

[0888] The "emotion engine" is an engine that recognizes the user's emotional state and reflects that information in the system's operation.

[0889] "Server" means the computer system that manages and processes data and fact-checking requests submitted by users.

[0890] A "collaborator" is an individual or organization whose role is to receive a fact-check request, verify its contents, and return the results.

[0891] A "queue" is a data structure for lining up received fact-check requests in order and waiting.

[0892] "Notification" means the act or means of communicating fact-check results or other information to a user.

[0893] "Training data" is data used to train and improve AI models.

[0894] The "reward system" is a mechanism that rewards contributors for providing excellent fact-checking results.

[0895] "Generative AI" is artificial intelligence that allows users to generate text.

[0896] A "terminal" is a device through which a user accesses the generative AI system.

[0897] The present invention provides a system for accessing a generation AI system on a user's terminal and verifying the authenticity of generated text. Specific embodiments for carrying out the present invention will be described below.

[0898] Overall system configuration

[0899] The system consists of the following main components:

[0900] 1. User devices: smartphones, tablets, computers, etc.

[0901] 2. Server: Manages the queue of fact-checking requests and stores the results.

[0902] 3. Generative AI: An AI engine that allows users to generate text.

[0903] 4. Emotion engine: Recognizes the user's emotional state and controls the system behavior.

[0904] 5. Fact-checking contributors: These contributors are responsible for verifying the authenticity of the generated text.

[0905] Program Overview

[0906] 1. Text Generation and Display

[0907] Users can use generative AI through their device to generate specific text, which is then instantly displayed on the device.

[0908] 2. Recognizing emotional states

[0909] The emotion engine uses the device's camera, microphone, and sensors to recognize the user's emotional state in real time, for example, using facial recognition and voice analysis technologies.

[0910] 3. Fact-checking requests

[0911] A button to request a fact check will be displayed next to the generated text. If the user has any concerns or doubts, they can press this button to send a fact check request to the server.

[0912] 4. Queue Management

[0913] The server adds the received fact-check request to a queue, and at the same time, the emotional state data sent from the emotion engine is also recorded on the server.

[0914] 5. Fact Check

[0915] Contributors access the server using a dedicated interface to retrieve fact-check requests added to the queue, check their contents, and send the results to the server.

[0916] 6. Notification of Results

[0917] The server notifies the user of the fact-check results received from the contributors and displays the results on the device. The results and data on the user's emotional state are also fed back as learning data for the AI ​​model, contributing to improving the accuracy of the system.

[0918] Hardware and Software

[0919] Key hardware and software used:

[0920] Smartphone: Camera and microphone for user interface and emotion recognition.

[0921] Server: Data management and fact-checking queue management.

[0922] Emotion Recognition API: Recognize user emotions in real time.

[0923] Fact Check API: Request fact checks and get results.

[0924] Notifications API: Notify users of fact-check results.

[0925] Specific examples

[0926] For example, suppose a user generates text such as "The highest mountain in the world is Mount Fuji" and has doubts while reading it. At this point, the emotion engine detects the user's emotions, and the system automatically requests a fact-check. The fact-check result, "The highest mountain in the world is Mount Everest," is provided and notified to the user's device.

[0927] Prompt Sentence Examples

[0928] text

[0929] If a user reads the article "Mount Fuji is the tallest mountain in the world" and has doubts, create a Python program to request a fact-check and display the results. Also, incorporate an engine to recognize the user's emotional state.

[0930] In this way, the present invention is a system that enables users to safely verify the authenticity of generated text and obtain highly reliable information.

[0931] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0932] Step 1:

[0933] A user accesses a generative AI system to generate text. The input is the content of the text the user wants to generate and a prompt, and the output is the generated text. The generative AI model generates text based on the prompt and displays it on the user's device.

[0934] Step 2:

[0935] The terminal displays the generated text and a button to request a fact check. This button is used when the user wants to verify the authenticity of the text. The input is the generated text and the state of the fact check button, and the output is the display of the button.

[0936] Step 3:

[0937] When a user presses the fact-check request button, the device sends the generated text to the server. At the same time, the emotion engine analyzes the user's emotional state and sends that information to the server. The input is the user's operation (button press) and emotional state data, and the output is data sent to the server.

[0938] Step 4:

[0939] The server adds the received generated text and emotional state information to a queue as a fact-checking request. At this time, the queue also contains metadata such as the date and time of the request's receipt and the emotional state. The input is the received data (generated text and emotional information), and the output is the fact-checking request added to the queue.

[0940] Step 5:

[0941] Fact-checking contributors access the server through a dedicated interface and retrieve fact-checking requests from the queue. The contributors check the contents of the generated text and conduct research to verify its authenticity. The input is the generated text retrieved from the queue, and the output is the fact-checking results.

[0942] Step 6:

[0943] The server receives the fact-check results returned by the contributors and stores them in a database. The saved fact-check results are displayed the next time the user accesses the server. The input is the fact-check results from the contributors, and the output is saving them in the database and notifying the user.

[0944] Step 7:

[0945] The server then feeds back the stored fact-check results and emotional state information as training data for the AI ​​model. This feedback improves the accuracy and reliability of the generative AI. The input is the fact-check results and emotional state information, and the output is an update to the AI ​​model.

[0946] Step 8:

[0947] The device receives a notification from the server and displays the generated fact-check result for the text, allowing the user to verify the authenticity of the information. The input is the fact-check result notification from the server, and the output is the result displayed on the user's device.

[0948] Through these steps, the system allows users to verify the authenticity of the generated text and realizes flexible fact-checking that takes into account emotional states.

[0949] 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.

[0950] 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.

[0951] 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.

[0952] [Fourth embodiment]

[0953] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0954] 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.

[0955] 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).

[0956] 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.

[0957] 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.

[0958] 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).

[0959] 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.

[0960] 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.

[0961] 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.

[0962] 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.

[0963] 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.

[0964] 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.

[0965] 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."

[0966] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[0967] First, a user accesses the generative AI system using their own device and generates text. The generated text is immediately displayed on the user's device, with a button next to the text to request fact-checking.

[0968] When a user presses this button, the corresponding text is sent to the server. The server adds the received text to a queue as a fact-check request. After the fact-check request is added to the queue, a contributor retrieves the request from the queue and works to verify the authenticity of the text. The contributor then sends the verification result to the server, which receives and stores the result.

[0969] The server then notifies the user of the received fact-check results. Specifically, the results are sent to the user's device and displayed on the user interface. The fact-check results are also used as training data for the AI ​​model. The server feeds the results back into the AI ​​model to improve its accuracy and reliability.

[0970] Furthermore, contributors who provide excellent fact-checking results will be rewarded. This reward system is expected to increase contributors' motivation and lead to higher quality fact-checking.

[0971] As a specific example, consider the case where a user generates text such as "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to a queue. A collaborator retrieves this request from the queue and obtains the check result, "Mount Everest is the highest." The result is sent to the server, which notifies the user of the result and provides feedback to the AI ​​model.

[0972] Each process in this system is carried out as follows: The user checks the text generated by the generative AI on their own device and sends a request to the server via the fact-check button. The server adds the request to a queue, receives and stores the results from contributors, and notifies the user. Finally, feedback is provided to the AI ​​model and contributors are rewarded. This series of steps allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole.

[0973] The above is an embodiment of the present invention. This system allows users to verify the authenticity of text without any burden, and enables them to quickly obtain accurate information.

[0974] The processing flow will be explained below.

[0975] Step 1:

[0976] A user uses a generative AI system to request text generation. Based on this request, the server instructs the generative AI to generate text. The generated text is sent from the server to the user's device and displayed on the device's screen.

[0977] Step 2:

[0978] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the generated text.

[0979] Step 3:

[0980] The user clicks the "Outsource fact-checking" button. The device takes the generated text and sends a fact-checking request in JSON format to the server.

[0981] Step 4:

[0982] The server processes the received fact-check request by adding it to a queue and notifying the user that the request has been received.

[0983] Step 5:

[0984] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor uses their own resources and knowledge to verify the authenticity of the generated text.

[0985] Step 6:

[0986] The collaborators compile the verification results and send them to the server, which receives them and stores them in a database.

[0987] Step 7:

[0988] The server notifies the user of the saved fact-check results, which are then displayed on the user's device, allowing the user to view the results.

[0989] Step 8:

[0990] The server feeds the fact-checking results back into the AI ​​model, which then undergoes a learning process to improve the model's accuracy and reliability.

[0991] Step 9:

[0992] The server evaluates the quality of the fact-check results provided by contributors and rewards contributors who provide excellent results. The server manages the distribution of rewards and notifies contributors.

[0993] This allows users to efficiently verify the authenticity of generated text, improving the reliability of the entire system.

[0994] Example 1

[0995] 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."

[0996] Currently, many generative AI systems generate text for users, but the lack of a means to verify the authenticity of the generated text has led to the spread of misinformation. Furthermore, the fact-checking process is often done manually, which is time-consuming and the quality of the checks cannot be guaranteed. Furthermore, there is a lack of mechanisms to maintain the motivation of contributors and continuously conduct high-quality fact-checks. Therefore, there is a need for the development of a system that can quickly and accurately verify the authenticity of generated text.

[0997] 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.

[0998] In this invention, the server includes: means for displaying the generated text so that the user can verify the authenticity of the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a contributor to obtain the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the artificial intelligence; means for displaying the result on the user's terminal based on the generated text; and means for calculating and notifying a reward to the contributor. This allows the user to quickly and accurately verify the authenticity of the generated text, improving the reliability of the entire system.

[0999] A "user terminal" is a device used by a user to display the generated text and request fact-checking.

[1000] "Generative AI" is artificial intelligence that generates text in response to a prompt.

[1001] The "fact check button" is an operation unit that allows the user to request confirmation of the authenticity of the generated text.

[1002] The "server" is a device that receives fact-checking requests, adds them to a queue, notifies the user of the results, and provides feedback to the artificial intelligence.

[1003] A "queue" is a data structure that allows a server to manage fact-check requests sequentially.

[1004] A "collaborator" is an individual or organization that responds to a fact-check request and verifies the authenticity of the text.

[1005] "Fact-check results" are information about the authenticity of a text that has been confirmed by a contributor.

[1006] "Artificial intelligence training data" refers to fact-checking results data used to improve the accuracy and reliability of generative AI.

[1007] "Rewards" are compensation paid to contributors who provide excellent fact-checking results.

[1008] The present invention provides a system for outsourcing fact-checking to verify the authenticity of the content of text output by a generative AI. Hereinafter, an embodiment of the present invention will be described in detail.

[1009] First, a user accesses the generative AI system using their own device and generates text. Specifically, they access the generative AI system's interface using a web browser. The user inputs a prompt sentence. For example, they might input "Mount Fuji is the highest mountain in the world." The generative AI model (e.g., OpenAI's GPT-3) generates text based on this prompt sentence. The generated text is immediately displayed on the user's device.

[1010] Next to the generated text is a button for requesting a fact check. When a user clicks this fact check button, the text is sent to the server as a fact check request. The server processes the received fact check request and adds it to a queue. This queue is a data structure for managing the processing order of requests.

[1011] The collaborator retrieves a request from the queue using a dedicated interface. The collaborator verifies the authenticity of the text content, for example, using the Internet or literature. For example, for the text "Mount Fuji is the highest mountain in the world," the collaborator obtains the result "Mount Everest is the highest." The verification result is entered into a dedicated input form and sent to the server.

[1012] The server stores the received fact-check results and notifies the user. The results are displayed on the user's device. For example, the result "Everest is the highest mountain" is displayed. The server also stores the fact-check results in a database for feedback to the generative AI model. This feedback improves the accuracy and reliability of the generative AI model.

[1013] In addition, contributors who provide excellent fact-checking results are rewarded. This reward can be paid via PayPal, for example. The server calculates the reward for contributors and notifies them.

[1014] As a concrete example of this system, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text is sent to the server and added to the queue. A collaborator retrieves this request from the queue and obtains the check result "Mount Everest is the highest." The result is sent to the server, which notifies the user and provides feedback to the generative AI model.

[1015] This process allows users to easily verify the authenticity of the generated text, improving the reliability of the system as a whole. Furthermore, by providing appropriate rewards to contributors, it is expected that high-quality fact-checking will be carried out sustainably. This concludes the description of the present invention.

[1016] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1017] Step 1:

[1018] A user accesses a generative AI system and inputs a prompt sentence. For example, they access the system's interface using a web browser. They input the prompt sentence, "Mount Fuji is the highest mountain in the world." Based on this input, a generative AI model (e.g., OpenAI's GPT-3) generates text. The generated text is displayed on the user's device screen. The input here is the prompt sentence, and the output is the generated text.

[1019] Step 2:

[1020] The user clicks on the fact-check button located next to the generated text. This click becomes the input, and a fact-check request is sent to the server along with the generated text. The server processes the received text as a fact-check request and adds it to a queue, which is a data structure. Here, the input is the user's click and the generated text, and the output is the server adding the fact-check request to the queue.

[1021] Step 3:

[1022] The server adds fact-check requests to a queue and monitors their status. Contributors retrieve requests from the queue using a dedicated interface. The retrieved text is sent to the contributor's device, and the contributor begins work to verify the text's content. The input here is the fact-check request added to the queue, and the output is the text received on the contributor's device.

[1023] Step 4:

[1024] A collaborator verifies the authenticity of the text content using the Internet or literature. For example, for a text that reads "Mount Fuji is the highest mountain in the world," the verification result is "Mount Everest is the highest mountain." The collaborator enters the verification result into a dedicated input form and sends it to the server. The input here is the text content and verification materials, and the output is the verification result.

[1025] Step 5:

[1026] The server stores the received confirmation result. The server notifies the user's terminal of this result. For example, the result "Everest is the highest mountain" is displayed on the user's terminal. The input here is the confirmation result, and the output is the notification to the user's terminal.

[1027] Step 6:

[1028] The server feeds back the fact-checking results to the generative AI model. The received results are stored in a database and used as training data for the generative AI model. The input here is the stored fact-checking results, and the output is reflected in the training data for the generative AI model.

[1029] Step 7:

[1030] The server calculates rewards for contributors. Contributors who provide good fact-check results are paid, for example, via PayPal. The payment of rewards includes a notification from the server to the contributor. The input here is the evaluation of the fact-check result, and the output is the payment of rewards to contributors.

[1031] (Application example 1)

[1032] 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."

[1033] The challenge is to quickly verify the reliability of text and product information generated by generative AI models, ensuring that users can safely use accurate information. It is also necessary to ensure the authenticity of product information provided in virtual stores, thereby improving users' shopping experiences.

[1034] 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.

[1035] In this invention, the server includes: means for a user to verify the authenticity of generated text; means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a fact-check request is made by the user; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, verify the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for an AI; means for providing product information to the user in a virtual store; means for displaying a button in the virtual store for requesting a fact-check on the generated product information; and means for requesting a fact-check through the product information displayed on the user's smart device. This allows users to quickly verify the authenticity of product information provided in the virtual store and enjoy shopping with peace of mind.

[1036] "User" refers to a person who uses a generative AI model to generate text or product information and verify the authenticity of that content.

[1037] "Generated text" refers to sentences or descriptions automatically created by a generative AI model.

[1038] "Server" refers to the computer system that manages the generated text and fact-checking requests, processes the fact-checking results from contributors, and notifies the user.

[1039] "Fact checking" refers to the process of verifying the authenticity of the content of generated text or product information.

[1040] A "collaborator" is someone who receives a fact-check request sent from the server, checks its contents, and sends the results to the server.

[1041] A "queue" refers to a data structure that stores received fact-check requests in order and provides them to contributors.

[1042] "AI training data" refers to data used to improve the accuracy and reliability of generative AI models.

[1043] "Virtual store" refers to a virtual store environment where products are sold and purchased over the Internet.

[1044] "Product information" refers to information such as product features, specifications, reviews, etc. presented in the virtual store.

[1045] "Smart devices" refers to portable electronic devices that can connect to the Internet, such as smartphones and smart glasses.

[1046] The present invention provides a system for outsourcing fact-checking to verify the authenticity of text and product information output by a generative AI model. The following describes in detail an embodiment of the present invention.

[1047] System Overview

[1048] The system consists of the following elements:

[1049] 1. User Device

[1050] 2. Cloud Server

[1051] 3. Collaborator terminal

[1052] 4. Generative AI Models

[1053] 5. Smart devices (smartphones, smart glasses)

[1054] System Operation

[1055] User device:

[1056] The user terminal provides a user interface for generating text and product information. The user inputs a prompt sentence into the generative AI model to generate the text and product information. The generated text and product information are displayed on the screen of the user terminal.

[1057] Cloud Server:

[1058] The cloud server plays a central role in managing fact-checking requests. When a user submits a fact-checking request, it is sent to the server and added to a queue. The server then provides the fact-checking requests to contributors and collects the results.

[1059] Collaborator's device:

[1060] The contributor terminal provides an interface for receiving fact-check requests. The contributor processes the requests and sends the results to the cloud server. Contributors are rewarded for providing good fact-check answers.

[1061] Generative AI models:

[1062] A generative AI model (e.g., GPT-4) generates text and product information based on a prompt entered by the user. This model runs on a cloud server and provides information in response to requests from user devices.

[1063] Smart Devices:

[1064] Users use smart devices (e.g., smartphones or smart glasses) to view the generated text and product information and submit fact-check requests.

[1065] Data processing and calculation

[1066] The cloud server processes data using the following software and hardware:

[1067] software:

[1068] Generative AI models (e.g., GPT-4)

[1069] Virtual store platforms (e.g., Shopify, WooCommerce)

[1070] Task management systems (e.g., Trello)

[1071] Hardware:

[1072] Cloud-based servers (e.g., Amazon Web Services) that provide computing power

[1073] Detailed procedure

[1074] 1. Information generation and display:

[1075] When a user accesses a product page in the virtual store, the generative AI model generates product information and displays it on the user's smart device.

[1076] 2. Fact-checking requests:

[1077] When a user presses the fact-check button, the generated text and product information are sent to the cloud server and added to the queue as a fact-check request.

[1078] 3. Collaborator verification:

[1079] The collaborator retrieves the request from the queue and checks the authenticity of the product information. The result is sent to the cloud server and notified to the user.

[1080] 4. Feedback of results:

[1081] The cloud server feeds the received results back into the AI ​​model to improve the model's accuracy.

[1082] Specific examples

[1083] For example, imagine a user wearing smart glasses accesses a virtual store and browses the product page for the latest smartphone. At this time, the generative AI model generates information such as "The latest smartphones are 5G compatible, have high-resolution cameras, and have long-lasting batteries." If the user wants to verify the authenticity of this information, they can click the fact-check button. The results of the check by the collaborators are then notified, and the user can receive reliable information such as "This smartphone is actually 5G compatible and has a long-lasting battery."

[1084] Prompt Sentence Examples

[1085] Generate user-specified product features and specifications.

[1086] Explain the effect of the fact-check button on product information.

[1087] This system allows users to quickly verify the authenticity of product information provided in virtual stores, allowing them to enjoy shopping with peace of mind.

[1088] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1089] Step 1:

[1090] The user device sends a request to the generative AI model.

[1091] Specifically, the user enters a prompt sentence and clicks the generate button. This sends the prompt sentence to the cloud server as an API request. The input data is the prompt sentence, and the output data is the generated text. The server invokes the generative AI model, generates text based on the entered prompt sentence, and sends the result to the user's device.

[1092] Step 2:

[1093] The user terminal displays the generated text.

[1094] The generated text received from the server is displayed on the user terminal. The user checks this text and determines its authenticity. The input data is the generated text received from the server, and the output data is the text displayed on the user interface.

[1095] Step 3:

[1096] The user terminal sends a fact-check request to the server.

[1097] When a user clicks the fact-check button, the generated text is sent to the cloud server as a fact-check request. The input data is the generated text and the fact-check request action, and the output data is the data to be added to the queue as a fact-check request. The server adds the request to a queue data structure.

[1098] Step 4:

[1099] The server provides fact-check requests to contributors.

[1100] The server retrieves fact-check requests from the queue and provides them to the collaborator terminal. The input data is the fact-check request in the queue, and the output data is the request sent to the collaborator terminal. The collaborator receives the request and starts checking it.

[1101] Step 5:

[1102] Contributors submit fact-check results to the server.

[1103] The contributor processes the request and sends the results to the cloud server. The input data is the fact-check request, and the output data is the verification result. The server stores the results received from the contributor and prepares them for further processing.

[1104] Step 6:

[1105] The server notifies the user terminal of the fact-check results.

[1106] The server sends the received fact-check results to the user's device. The input data is the confirmation result received from the collaborator, and the output data is a notification displayed on the user's device. The user can check the result.

[1107] Step 7:

[1108] The server feeds the fact-checking results back to the generative AI model.

[1109] The server feeds back the fact-check results as training data for the generative AI model. This feedback improves the accuracy and reliability of the AI ​​model. The input data is the fact-check results, and the output data is an updated AI model.

[1110] Step 8:

[1111] The user's smart device displays the product information.

[1112] The generated product information is displayed on the smart device, allowing for fact-checking requests. The input data is the generated product information, and the output data is the information displayed on the smart device. The user can click to submit the request.

[1113] 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.

[1114] The present invention is a system that outsources fact-checking to verify the authenticity of the content of text output by generative AI, and further combines it with an emotion engine that recognizes the emotional state of the user and controls the operation of the system based on that. Below, we will explain in detail the embodiments of the present invention.

[1115] First, the user accesses the generative AI system using their own device and generates text. This generated text is immediately displayed on the user's device. Next to the generated text, there is a button to request a fact check. This button is used if the user wants to verify the authenticity of the generated text.

[1116] Next, when the user presses this button, the corresponding text is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received text and emotional state information to a queue as a fact-check request and notifies the user of the receipt of the request.

[1117] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated text and sends the results to the server. The server receives the results and stores them in a database. The saved fact-check results are notified to the user and displayed on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[1118] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[1119] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[1120] As a specific example, consider the case where a user generates the text "Mount Fuji is the highest mountain in the world" and requests that it be fact-checked. This text and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check result, and the server notifies the user of the result. At the same time, the result and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[1121] The above is an embodiment of the present invention. The introduction of an emotion engine enables flexible fact-checking that takes into account the emotional state of the user, further improving the usability and reliability of the system.

[1122] The processing flow will be explained below.

[1123] This invention is a system that outsources fact-checking to verify the authenticity of the text content output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that. The processing flow is explained below in specific steps.

[1124] Step 1:

[1125] A user accesses the generative AI system using a device and requests text generation. The server instructs the generative AI to generate text and sends the generated text to the user's device. The device then displays the generated text on its screen.

[1126] Step 2:

[1127] The device will display the generated text along with a "Outsource fact-check" button, which can be used if the user wants to verify the authenticity of the text.

[1128] Step 3:

[1129] The user clicks the "Outsource fact-checking" button. The device retrieves the generated text, and the emotion engine recognizes the user's current emotional state. This information is then sent to the server in JSON format as a fact-checking request.

[1130] Step 4:

[1131] The server adds the received fact-check request to a queue and notifies the user with a message confirming receipt of the request.

[1132] Step 5:

[1133] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor verifies the authenticity of the generated text and sends the result to the server.

[1134] Step 6:

[1135] The server receives fact-check results from contributors, stores them in a database, and notifies the user of the results and displays them on the user's device.

[1136] Step 7:

[1137] The server then feeds back the stored fact-check results and the user's emotional state information to the AI ​​model, which is used to improve the model's performance.

[1138] Step 8:

[1139] The emotion engine determines the need for fact-checking based on the user's emotional state. If the user expresses anxiety or doubt, the system automatically sends a fact-check request to the server, reducing the burden on the user.

[1140] Step 9:

[1141] The server evaluates the quality of the fact-check results provided by contributors, rewards contributors who provide excellent results, and manages the distribution of rewards and notifies contributors.

[1142] Through these steps, users can efficiently verify the authenticity of the generated text, improving the reliability and usability of the entire system. The introduction of an emotion engine enables flexible responses that take into account the user's psychological state, quickly ensuring the accuracy of information.

[1143] Example 2

[1144] 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."

[1145] There is a need for a method to quickly and accurately verify the authenticity of user-generated documents. Conventional systems have problems: fact-checking takes time and it is difficult to respond appropriately taking into account the user's emotional state. This can lead to users becoming anxious about the reliability of the information, which can lead to a decrease in the reliability of the system.

[1146] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for displaying a document created by a user, means for providing a button for requesting a fact-check for the document, means for transmitting the document to the server when a fact-check request is made by the user, means for adding the fact-check request received by the server to a queue, means for a collaborator to acquire the fact-check request from the queue, check the content, and transmit the result to the server, means for notifying the user of the fact-check result received by the server, means for feeding back the result and information on the user's emotional state as learning data for the AI, and an emotion engine that recognizes the user's emotional state and automatically requests a fact-check. This makes it possible to quickly and accurately verify the authenticity of a document while taking the user's emotional state into consideration.

[1147] A "user" is an entity that uses the system to generate documents and request fact-checking.

[1148] "Document" refers to text data created by generative AI.

[1149] "Generative AI" is an artificial intelligence model that generates documents based on prompts entered by the user.

[1150] "Fact checking" is a verification process to confirm the authenticity of the contents of a document.

[1151] A "button" is an interface element that a user clicks on the screen to request a specific action from the system.

[1152] A "server" is a computer system that processes and stores user requests and data.

[1153] A "queue" is a data structure for queuing received fact-check requests.

[1154] A "collaborator" is a person or system that receives a fact-check request, verifies its contents, and provides the results.

[1155] "Emotional state information" is data that represents the emotional state of the user.

[1156] "Training data" is a collection of past data and results used to improve the performance of an AI model.

[1157] An "emotion engine" is software or hardware that recognizes the user's emotional state and controls the system's behavior based on that.

[1158] "Feedback" refers to the re-input and re-processing of specific data or results for the purpose of improving the system.

[1159] "Notification" is a communication method by which a server provides information to a user.

[1160] A "prompt" refers to an instruction or question that is input to a generative AI.

[1161] This invention is a system that outsources fact-checking to verify the authenticity of the contents of documents output by generative AI, and also combines it with an emotion engine that recognizes the user's emotional state and controls the system's operation based on that.

[1162] First, a user uses their own device to access a generative AI system (e.g., a generative AI model) and enters a prompt to generate a document. The generated document is immediately displayed on the user's device. Next to the generated document, there is a button for requesting fact-checking. This button is used when the user wants to verify the authenticity of the generated document.

[1163] Next, when the user presses the fact-check button, the corresponding document is sent to the server. At the same time, the emotion engine recognizes the user's current emotional state and sends that information to the server. The server adds the received document and emotional state information to a queue (e.g., AWS SQS) as a fact-check request and notifies the user of the receipt of the request.

[1164] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. The contributor checks the authenticity of the generated document and sends the results to the server. The server receives the results and stores them in a database (e.g., MySQL). The server notifies the user of the stored fact-check results and displays them on the user's device. In addition, the fact-check results and the user's emotional state information are fed back as learning data for the AI ​​model. This makes it possible to improve the accuracy and reliability of the model.

[1165] Furthermore, the system can determine the need for fact-checking based on the user's emotional state as recognized by the emotion engine and automatically request a fact-check. For example, if the user is clearly feeling anxious or suspicious, the system can automatically request a fact-check based on that emotional state. This eliminates the need for the user to manually request a fact-check, allowing the system to quickly and appropriately verify the authenticity of information.

[1166] In addition, the quality of the fact-check results provided by contributors can be evaluated, and contributors who provide excellent results can be rewarded. This will increase contributors' motivation and promote high-quality fact-checking. This reward system is expected to encourage active participation from contributors.

[1167] As a specific example, consider the case where a user generates a document stating "Mount Fuji is the tallest mountain in the world" and requests that it be fact-checked. This document and the user's emotional state information (e.g., emotion indicating doubt) are sent to the server. The server adds this information to a queue and sends the request to a collaborator. The collaborator returns the fact-check results, and the server notifies the user of the results. At the same time, the results and emotional state information are fed back to the AI ​​model, improving the accuracy of the system.

[1168] The introduction of an emotion engine enables flexible fact-checking that takes into account the user's emotional state, further improving the usability and reliability of the system.

[1169] An example of a prompt sentence is "Generate the document 'Mount Fuji is the highest mountain in the world' and request a fact check."

[1170] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1171] Step 1:

[1172] A user accesses the generative AI system using their own device. The user inputs a prompt and requests text generation. The device sends the user's input (prompt) to the generative AI model and receives the generated text. For example, if a user inputs the prompt "Tell me about Mount Fuji," the device receives the text "Mount Fuji is the highest mountain in Japan" from the generative AI. This generated text is immediately displayed on the device.

[1173] Step 2:

[1174] The user presses a fact-check button displayed next to the generated text. The device sends the generated text and the user's emotional state (e.g., anxiety or doubt) to the server. The input here is the generated text and the user's emotional state information, which the device then sends to the server. For example, the device sends the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state information of anxiety.

[1175] Step 3:

[1176] The server adds the received text and emotional state information to a queue. As input, the server receives the generated text and emotional state information received from the device. The server adds this information to a queue (e.g., AWS SQS) and processes it as a fact-checking request. For example, the text "Mount Fuji is the highest mountain in Japan" and the user's emotional state are added to the queue.

[1177] Step 4:

[1178] The server notifies the user that it has received the fact-checking request. The server confirms that the fact-checking request has been added to the queue and notifies the user that it has been received. The input is the queue information added to the server, and the output is a notification to the user. For example, a notification saying "your fact-checking request has been accepted" is sent to the user.

[1179] Step 5:

[1180] A contributor accesses the fact-checking system and retrieves a fact-check request from the queue. As input, the contributor retrieves a fact-check request from the queue, and as output, the retrieved text information is obtained. For example, a contributor retrieves the text "Mount Fuji is the highest mountain in Japan."

[1181] Step 6:

[1182] The collaborator verifies the truth of the generated text content and sends the result to the server. As input, the collaborator receives the text information to be verified and investigates its content. As output, the investigation result is sent to the server. For example, the collaborator verifies the result that "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" and sends it to the server.

[1183] Step 7:

[1184] The server receives the fact-check results and stores them in a database. As input, it receives the fact-check results sent by the contributors, and as output, it stores them in a database. For example, the result "Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world" is stored in the database.

[1185] Step 8:

[1186] The server notifies the user of the saved fact-check results and displays the results on the user's device. As input, the server retrieves the saved results from the database and notifies the user as output. The user can view the fact-check results on their device. For example, a notification is displayed to the user saying, "Fact-check result: Mount Fuji is the highest mountain in Japan, but not the highest mountain in the world."

[1187] Step 9:

[1188] The server feeds back the fact-check results and the user's emotional state information as learning data for the AI ​​model. The fact-check results and emotional state information are used as input, and data to improve the accuracy of the AI ​​model is obtained as output. For example, the text "Mount Fuji is the highest mountain in Japan," the emotional state "anxiety," and the fact-check results are fed back to the AI ​​model.

[1189] Step 10:

[1190] The emotion engine recognizes the user's emotional state and automatically requests fact-checking when necessary. The user's emotional state information is used as input, and an automatic fact-checking request is generated as output. For example, if the user expresses anxiety, the system will automatically request a fact-check.

[1191] By using the above steps, the system of the present invention can quickly and accurately verify the authenticity of a document while taking into account the emotional state of the user.

[1192] (Application example 2)

[1193] 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."

[1194] In recent years, text generation using generative AI has become widespread, and the importance of fact-checking to verify the authenticity of generated text is increasing. However, the process for users to verify the reliability of generated text is time-consuming and often leaves users feeling anxious or suspicious. Furthermore, a system that takes into account the user's emotional state is needed to efficiently verify the authenticity of generated text. Furthermore, a reward system for contributors is also necessary to improve the quality of fact-checking results. The present invention aims to provide a system that solves these problems.

[1195] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for displaying the generated text; means for providing a button for requesting a fact-check on the text; means for transmitting the text to the server when a user requests a fact-check; means for adding the fact-check request received by the server to a queue; means for a collaborator to retrieve the fact-check request from the queue, check the content, and transmit the result to the server; means for notifying the user of the fact-check result received by the server; means for feeding back the result as learning data for the AI; means for recognizing the user's emotional state using an emotion engine and controlling the system operation based on that information; means for automatically requesting a fact-check according to the emotional state; and means for displaying the fact-check result based on the emotional state on the user's terminal. This allows the user to quickly and efficiently verify the authenticity of the generated text, thereby realizing flexible fact-checking that takes the user's emotional state into consideration.

[1196] "User" means an individual or organization that uses the generative AI system.

[1197] "Generated text" refers to textual information generated by a generation AI.

[1198] "Fact checking" is the act of verifying the truth of the content of the generated text.

[1199] The "emotion engine" is an engine that recognizes the user's emotional state and reflects that information in the system's operation.

[1200] "Server" means the computer system that manages and processes data and fact-checking requests submitted by users.

[1201] A "collaborator" is an individual or organization whose role is to receive a fact-check request, verify its contents, and return the results.

[1202] A "queue" is a data structure for lining up received fact-check requests in order and waiting.

[1203] "Notification" means the act or means of communicating fact-check results or other information to a user.

[1204] "Training data" is data used to train and improve AI models.

[1205] The "reward system" is a mechanism that rewards contributors for providing excellent fact-checking results.

[1206] "Generative AI" is artificial intelligence that allows users to generate text.

[1207] A "terminal" is a device through which a user accesses the generative AI system.

[1208] The present invention provides a system for accessing a generation AI system on a user's terminal and verifying the authenticity of generated text. Specific embodiments for carrying out the present invention will be described below.

[1209] Overall system configuration

[1210] The system consists of the following main components:

[1211] 1. User devices: smartphones, tablets, computers, etc.

[1212] 2. Server: Manages the queue of fact-checking requests and stores the results.

[1213] 3. Generative AI: An AI engine that allows users to generate text.

[1214] 4. Emotion engine: Recognizes the user's emotional state and controls the system behavior.

[1215] 5. Fact-checking contributors: These contributors are responsible for verifying the authenticity of the generated text.

[1216] Program Overview

[1217] 1. Text Generation and Display

[1218] Users can use generative AI through their device to generate specific text, which is then instantly displayed on the device.

[1219] 2. Recognizing emotional states

[1220] The emotion engine uses the device's camera, microphone, and sensors to recognize the user's emotional state in real time, for example, using facial recognition and voice analysis technologies.

[1221] 3. Fact-checking requests

[1222] A button to request a fact check will be displayed next to the generated text. If the user has any concerns or doubts, they can press this button to send a fact check request to the server.

[1223] 4. Queue Management

[1224] The server adds the received fact-check request to a queue, and at the same time, the emotional state data sent from the emotion engine is also recorded on the server.

[1225] 5. Fact Check

[1226] Contributors access the server using a dedicated interface to retrieve fact-check requests added to the queue, check their contents, and send the results to the server.

[1227] 6. Notification of Results

[1228] The server notifies the user of the fact-check results received from the contributors and displays the results on the device. The results and data on the user's emotional state are also fed back as learning data for the AI ​​model, contributing to improving the accuracy of the system.

[1229] Hardware and Software

[1230] Key hardware and software used:

[1231] Smartphone: Camera and microphone for user interface and emotion recognition.

[1232] Server: Data management and fact-checking queue management.

[1233] Emotion Recognition API: Recognize user emotions in real time.

[1234] Fact Check API: Request fact checks and get results.

[1235] Notifications API: Notify users of fact-check results.

[1236] Specific examples

[1237] For example, suppose a user generates text such as "The highest mountain in the world is Mount Fuji" and has doubts while reading it. At this point, the emotion engine detects the user's emotions, and the system automatically requests a fact-check. The fact-check result, "The highest mountain in the world is Mount Everest," is provided and notified to the user's device.

[1238] Prompt Sentence Examples

[1239] text

[1240] If a user reads the article "Mount Fuji is the tallest mountain in the world" and has doubts, create a Python program to request a fact-check and display the results. Also, incorporate an engine to recognize the user's emotional state.

[1241] In this way, the present invention is a system that enables users to safely verify the authenticity of generated text and obtain highly reliable information.

[1242] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1243] Step 1:

[1244] A user accesses a generative AI system to generate text. The input is the content of the text the user wants to generate and a prompt, and the output is the generated text. The generative AI model generates text based on the prompt and displays it on the user's device.

[1245] Step 2:

[1246] The terminal displays the generated text and a button to request a fact check. This button is used when the user wants to verify the authenticity of the text. The input is the generated text and the state of the fact check button, and the output is the display of the button.

[1247] Step 3:

[1248] When a user presses the fact-check request button, the device sends the generated text to the server. At the same time, the emotion engine analyzes the user's emotional state and sends that information to the server. The input is the user's operation (button press) and emotional state data, and the output is data sent to the server.

[1249] Step 4:

[1250] The server adds the received generated text and emotional state information to a queue as a fact-checking request. At this time, the queue also contains metadata such as the date and time of the request's receipt and the emotional state. The input is the received data (generated text and emotional information), and the output is the fact-checking request added to the queue.

[1251] Step 5:

[1252] Fact-checking contributors access the server through a dedicated interface and retrieve fact-checking requests from the queue. The contributors check the contents of the generated text and conduct research to verify its authenticity. The input is the generated text retrieved from the queue, and the output is the fact-checking results.

[1253] Step 6:

[1254] The server receives the fact-check results returned by the contributors and stores them in a database. The saved fact-check results are displayed the next time the user accesses the server. The input is the fact-check results from the contributors, and the output is saving them in the database and notifying the user.

[1255] Step 7:

[1256] The server then feeds back the stored fact-check results and emotional state information as training data for the AI ​​model. This feedback improves the accuracy and reliability of the generative AI. The input is the fact-check results and emotional state information, and the output is an update to the AI ​​model.

[1257] Step 8:

[1258] The device receives a notification from the server and displays the generated fact-check result for the text, allowing the user to verify the authenticity of the information. The input is the fact-check result notification from the server, and the output is the result displayed on the user's device.

[1259] Through these steps, the system allows users to verify the authenticity of the generated text and realizes flexible fact-checking that takes into account emotional states.

[1260] 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.

[1261] 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.

[1262] 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 robot 414.

[1263] 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.

[1264] FIG. 9 illustrates 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 behaviors 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.

[1265] 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.

[1266] 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).

[1267] 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.

[1268] 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."

[1269] 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.

[1270] 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).

[1271] 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.

[1272] 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.

[1273] 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.

[1274] 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.

[1275] 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.

[1276] 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.

[1277] 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.

[1278] 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.

[1279] 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.

[1280] 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.

[1281] The following is further disclosed regarding the above embodiment.

[1282] (Claim 1)

[1283] For users to verify the authenticity of the generated text,

[1284] a means for displaying the generated text;

[1285] means for providing a button to request a fact check on the text;

[1286] means for transmitting the text to a server when a fact-check request is made by a user;

[1287] means for adding received fact-check requests to a queue;

[1288] A means for a collaborator to obtain a fact-check request from the queue, check the content, and send the result to the server;

[1289] means for notifying a user of the fact-check results received by the server;

[1290] A means for feeding back the results as learning data for the AI;

[1291] A system including:

[1292] (Claim 2)

[1293] 10. The system of claim 1, further comprising means for rewarding contributors for good fact-check answers.

[1294] (Claim 3)

[1295] 10. The system of claim 1, further comprising means for incorporating generative AI for enabling a user to generate text.

[1296]

[1297] "Example 1"

[1298] (Claim 1)

[1299] For users to verify the authenticity of the generated text,

[1300] a means for displaying the generated text;

[1301] means for providing a button to request a fact check on the text;

[1302] means for transmitting the text to a server when a fact-check request is made by a user;

[1303] means for adding received fact-check requests to a queue;

[1304] A means for a collaborator to obtain a fact-check request from the queue, check the content, and send the result to the server;

[1305] means for notifying a user of the fact-check results received by the server;

[1306] A means for feeding back the results as learning data for the artificial intelligence;

[1307] means for displaying results on a user's terminal based on the generated text;

[1308] A means for calculating and notifying rewards to contributors;

[1309] A system including:

[1310] (Claim 2)

[1311] 10. The system of claim 1, further comprising means for rewarding contributors for good fact-check answers.

[1312] (Claim 3)

[1313] 10. The system of claim 1, further comprising means for incorporating a generative artificial intelligence for engaging a user in generating text.

[1314] "Application Example 1"

[1315] (Claim 1)

[1316] a means for a user to verify the authenticity of the generated text;

[1317] a means for displaying the generated text;

[1318] means for providing a button to request a fact check on the text;

[1319] means for transmitting the text to a server when a fact-check request is made by a user;

[1320] means for adding received fact-check requests to a queue;

[1321] A means for a collaborator to obtain a fact-check request from the queue, check the content, and send the result to the server;

[1322] means for notifying a user of the fact-check results received by the server;

[1323] A means for feeding back the results as learning data for the AI;

[1324] means for providing product information to users within the virtual store;

[1325] a means for displaying a button in the virtual store to request a fact check on the generated product information;

[1326] A means for requesting fact-checking through product information displayed on a user's smart device;

[1327] A system including:

[1328] (Claim 2)

[1329] 10. The system of claim 1, further comprising means for rewarding contributors for good fact-check answers.

[1330] (Claim 3)

[1331] 10. The system of claim 1, further comprising means for incorporating generative AI for enabling a user to generate text.

[1332] "Example 2: Combining Emotion Engines"

[1333] (Claim 1)

[1334] For users to verify the authenticity of generated documents,

[1335] a means for displaying the generated document;

[1336] means for providing a button to request a fact check on the document;

[1337] a means for transmitting a document to a server when a fact-check request is made by a user;

[1338] means for adding received fact-check requests to a queue;

[1339] A means for a collaborator to obtain a fact-check request from the queue, check the content, and send the result to the server;

[1340] means for notifying a user of the fact-check results received by the server;

[1341] A means for feeding back the results and the user's emotional state information as learning data for the AI;

[1342] means for recognizing a user's emotional state and automatically making fact-check requests, the emotional engine comprising:

[1343] A system including:

[1344] (Claim 2)

[1345] 10. The system of claim 1, further comprising means for rewarding contributors for good fact-check answers.

[1346] (Claim 3)

[1347] 10. The system of claim 1, further comprising means for incorporating a generative AI for enabling a user to generate documents.

[1348] "Application example 2 when combining emotion engines"

[1349] (Claim 1)

[1350] For users to verify the authenticity of the generated text,

[1351] a means for displaying the generated text;

[1352] means for providing a button to request a fact check on the text;

[1353] means for transmitting the text to a server when a fact-check request is made by a user;

[1354] means for adding received fact-check requests to a queue;

[1355] A means for a collaborator to obtain a fact-check request from the queue, check the content, and send the result to the server;

[1356] means for notifying a user of the fact-check results received by the server;

[1357] A means for feeding back the results as learning data for the AI;

[1358] a means for recognizing the user's emotional state by an emotion engine and controlling the system's operation based on that information;

[1359] A means to automatically request fact-checks depending on emotional state;

[1360] means for displaying the results of a fact check based on the emotional state on a user's device;

[1361] A system including:

[1362] (Claim 2)

[1363] 10. The system of claim 1, further comprising means for rewarding contributors for good fact-check answers.

[1364] (Claim 3)

[1365] 10. The system of claim 1, further comprising means for incorporating generative AI for enabling a user to generate text. [Explanation of symbols]

[1366] 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. For users to verify the authenticity of the generated text, a means for displaying the generated text; means for providing a button to request a fact check on the text; means for transmitting the text to a server when a fact-check request is made by a user; means for adding received fact-check requests to a queue; A means for a collaborator to obtain a fact-check request from the queue, check the content, and send the result to the server; means for notifying a user of the fact-check results received by the server; A means for feeding back the results as learning data for the AI; A system including:

2. The system of claim 1 , further comprising means for rewarding contributors for good fact-check answers.

3. 10. The system of claim 1, further comprising means for incorporating a generative AI for allowing a user to generate text.

Images