system

Abstract

A system is provided.SOLUTION: Means for inputting user certification information and transmitting the user certification information to a server, means for the server to verify the certification information and generate a login token in a case where the user is certified, means for providing a quiz task to the certified user instead of advertisement watching, means for the user to answer the quiz and transmit the answer to the server, means for the server to receive answer information, remove inappropriate information, and add the answer information to a training dataset, and train a AI model on the basis of the training dataset, the system includes a means for generating the model as a plug-in, a means for providing the plug-in to companies in a specific field via an API, and a means for giving a reward each time a user completes a quiz task.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Generative AI faces the challenge of providing low-quality answers in specific fields due to a lack of training data in those fields. Furthermore, viewing advertisements is often tedious and distracting for application users, often detracting from the user experience. The present invention aims to improve the performance of generative AI and the user experience by having application users create training data for generative AI through quizzes instead of watching advertisements. [Means for solving the problem]

[0005] The present invention proposes a system that provides users with quiz-style tasks instead of viewing advertisements when using an application. The system includes a means for inputting user authentication information and transmitting it to a server, a means for the server to verify the authentication information and generate a login token if authenticated, a means for providing authenticated users with quiz tasks instead of viewing advertisements, a means for the user to answer the quiz and transmit the answers to the server, a means for the server to receive the answer data, remove inappropriate data, and add it to a training dataset, a means for training an AI model based on the training dataset and generating the model as a plugin, and a means for providing the plugin to companies in specific fields via an API. The system also includes a means for rewarding users each time they complete a quiz task, thereby maintaining user motivation. In this way, the present invention simultaneously improves the performance of the generation AI and the user experience.

[0006] "User authentication information" means information required to identify a user and authorize access, typically a combination of a username and password.

[0007] A "server" is a computer system that manages, stores, and provides data over a network.

[0008] A "token" is a character string that is issued by the server to a user who has been authorized to authenticate, and that serves to prove the validity of the session.

[0009] "Advertisement viewing" refers to the act of a user viewing a commercial advertisement displayed while using an application.

[0010] A "quiz task" is a question-type task that is provided in place of watching an advertisement, and is intended to test the user's knowledge.

[0011] A "user answer" is an answer entered by a user in response to a quiz task.

[0012] A "training dataset" is a collection of data that a generative AI uses to train its model.

[0013] An "AI model" is an artificial intelligence algorithm or software that has been trained using a training dataset.

[0014] "Plugins" are additional modules generated in the form of AI models to provide specific functions or services.

[0015] "API" stands for Application Program Interface, a standardized means of exchanging data and functions between different software programs.

[0016] "Rewards" refer to points or benefits awarded to users each time they complete a quiz task. [Brief explanation of the drawings]

[0017] [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

[0018] 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.

[0019] First, the terms used in the following description will be explained.

[0020] 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).

[0021] 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.

[0022] 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.

[0023] 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.

[0024] 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."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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."

[0038] The present invention is a system that provides quiz-style tasks in exchange for watching advertisements, and collects learning data for a generation AI by having users answer the quiz tasks, and trains an AI model based on this data. The following describes in natural language the roles and specific operations of the user, terminal, and server in an embodiment of the present invention.

[0039] 1. User authentication and login

[0040] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0041] The terminal transmits the entered user authentication information to the server.

[0042] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0043] The device will notify the user that login was successful and transition to the main screen.

[0044] 2. Submit the quiz task

[0045] After logging in, the server checks the usage history based on the user ID.

[0046] If the server determines that it is time to watch an ad, it generates a quiz-style learning task, such as "Which of the following products will be the best-selling in 2022?" with multiple choices.

[0047] The server sends the generated quiz task to the terminal.

[0048] The terminal notifies the user of the quiz task and displays a quiz screen.

[0049] 3. User responses and data submission

[0050] The user selects one of the options displayed (e.g., "smartphone," "tablet," "laptop," or "smartwatch") and presses the send button.

[0051] The terminal transmits the selected answer to the server.

[0052] The server temporarily stores the received response data and proceeds to the next process.

[0053] 4. Data processing and validation

[0054] The server analyzes the received response data and removes inappropriate data and outliers.

[0055] The remaining clean data is added to the AI ​​training dataset.

[0056] For example, answer data collected from multiple users to the question, "What was the best-selling product in 2022?" is integrated and provided to the generative AI as training data.

[0057] 5. Creating and Providing Plugins

[0058] The server trains the AI ​​model using the clean data.

[0059] A trained AI model can be plugged into a specific field (e.g., "product trend prediction").

[0060] The server provides the generated plugins to companies in specific fields or their own services via API.

[0061] 6. User Rewards and Feedback

[0062] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to their accounts.

[0063] The terminal displays a notification to the user that points have been earned.

[0064] Users can check the points and rewards they have earned, which will motivate them to take part in the next quiz task.

[0065] Specific examples

[0066] 1. User authentication and login details example

[0067] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[0068] The terminal sends this information to the server, which checks the information against a database.

[0069] If the authentication is successful, the server generates a token and returns it to the terminal.

[0070] The device will notify you that the login was successful and will transition to the main screen.

[0071] 2. Example of Quiz Task Submission Details

[0072] The server checks the usage history of user "alice" and determines that it is time to view an advertisement.

[0073] The server generates a quiz task: "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[0074] The terminal displays the quiz and notifies the user.

[0075] 3. Detailed example of user responses and data submission

[0076] The user selects "smartphone" and presses the send button.

[0077] The terminal transmits this response data to the server.

[0078] The server analyzes the received response data and removes inappropriate data.

[0079] 4. Detailed examples of data processing and validation

[0080] The server adds "smartphone" to the training dataset as clean data.

[0081] 5. Detailed example of creating and providing a plugin

[0082] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[0083] The server provides this plugin to companies and their own services via API.

[0084] 6. Detailed Example of User Rewards and Feedback

[0085] The server calculates a reward of 10 points for "alice" and adds it to her account.

[0086] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[0087] This allows users to participate in quizzes instead of watching ads, and also improves the performance of the generative AI.

[0088] The processing flow will be explained below.

[0089] Step 1:

[0090] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0091] Step 2:

[0092] The terminal transmits the entered user authentication information to the server.

[0093] Step 3:

[0094] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0095] Step 4:

[0096] The device will notify the user that login was successful and transition to the main screen.

[0097] Step 5:

[0098] The server checks the usage history based on the user ID and determines whether it is time to watch an advertisement.

[0099] Step 6:

[0100] If the server determines that an ad is required, it generates a quiz task. For example, it prepares a question such as, "Which of the following products was the best-selling in 2022?" and presents several options.

[0101] Step 7:

[0102] The server sends the generated quiz task to the terminal.

[0103] Step 8:

[0104] The terminal notifies the user of the quiz task and displays a quiz screen.

[0105] Step 9:

[0106] The user selects one of the options displayed on the quiz screen. For example, they select "Smartphone" and press the submit button.

[0107] Step 10:

[0108] The terminal transmits the selected answer to the server.

[0109] Step 11:

[0110] The server temporarily stores the received response data and starts data cleaning.

[0111] Step 12:

[0112] The server analyzes the received response data and removes inappropriate data and outliers.

[0113] Step 13:

[0114] The server adds the clean data to the AI's training dataset.

[0115] Step 14:

[0116] The server trains the AI ​​model based on the clean data.

[0117] Step 15:

[0118] The server generates a trained AI model as a generative AI plugin for a specific field.

[0119] Step 16:

[0120] The server provides this plugin to companies in specific fields or their own services via API.

[0121] Step 17:

[0122] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[0123] Step 18:

[0124] The terminal displays a notification to the user that points have been earned.

[0125] Example 1

[0126] 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."

[0127] Conventional ad-based services have the problem of being monotonous for users and low engagement. Furthermore, the revenue model based on ad viewing is unstable because it depends on the number of views. Furthermore, the accuracy of the collected data is low, making it difficult to obtain high-quality data, which is important for training generative AI models. A new system that can solve these issues and collect high-quality data while increasing user engagement is needed.

[0128] 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.

[0129] In this invention, the server includes means for inputting user authentication information and transmitting it from the terminal to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz-style task in exchange for watching an advertisement, means for the user to answer the quiz-style task and transmit the answer from the terminal to the server, means for the server to receive the answer data, analyze it to remove inappropriate data, and add the clean data to a training dataset, means for training a generative AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to organizations in a specific field via an API, and means for granting a reward each time a user completes a quiz-style task. This makes it possible to collect high-quality data and effectively train the generative AI model while increasing user engagement.

[0130] "User credentials" are the information a user enters to identify themselves and gain access, often consisting of a username and password.

[0131] A "terminal" is a device operated by a user to access the system, and includes, for example, a smartphone or a personal computer.

[0132] A "server" is a computer system that provides services over a network, and is a device that has the functions of data processing, storage, and responding to requests from clients.

[0133] A "login token" is authentication information that indicates that a user has been authenticated and is used to maintain a logged-in state for a certain period of time.

[0134] A "quiz-style task" is a quiz question that a user must answer, and is a task that includes multiple-choice questions and multiple options.

[0135] "Clean data" refers to high-quality data remaining after removing inappropriate data and outliers from the received response data.

[0136] A "training dataset" is a collection of data used to train a generative AI model, to which clean data has been added.

[0137] A "generative AI model" is a model of artificial intelligence that is trained based on collected data and algorithms used to perform specific tasks.

[0138] A "plug-in" is a software component with specific functionality that can be added to other software systems to extend their functionality.

[0139] "API" stands for Application Programming Interface, a standardized interface for communication between different software systems.

[0140] "Rewards" are points, special benefits, or other compensation that a user receives each time they complete a quiz-style task.

[0141] This invention is a system that collects high-quality data while increasing user engagement through a series of processes including user authentication, quiz-style task provision, data collection and analysis, generative AI model training, and a reward system. Below, we will explain the specific hardware and software usage methods and data processing and data calculation procedures.

[0142] User Authentication and Login

[0143] The device starts up, such as a smartphone or PC, and runs an application. The application displays a screen for entering user authentication information (username and password). The user enters this information and presses the login button. The device then sends the entered authentication information to the server using an HTTPS request.

[0144] The server checks the received authentication information against the database, and if authentication is successful, generates a JSON Web Token (JWT). This token is returned to the terminal, and a notification of successful login is displayed to the user, and the user is taken to the main screen.

[0145] Providing quiz-style tasks

[0146] After logging in, the server checks the user's usage history based on the user ID. If the usage history meets certain conditions, the server creates a quiz-style task instead of watching an advertisement. For example, the server generates a question such as "Which of the following products was the best-selling product in 2022?" along with options such as "smartphone," "tablet," "laptop," and "smartwatch."

[0147] The server sends this quiz task to the terminal, and the terminal notifies the user. The quiz screen is displayed on the terminal screen, and the user selects an answer from the displayed options.

[0148] User responses and data submission

[0149] The user selects an answer from the options and presses the submit button. The device sends the answer data to the server using an HTTPS request.

[0150] The server receives the response data and temporarily stores it. Next, it analyzes the data using Python's pandas library or similar to remove inappropriate data and outliers. The remaining clean data is added to the training dataset.

[0151] Training generative AI models

[0152] The server uses the clean data to train a generative AI model. For example, it trains a neural network model using frameworks such as TensorFlow or PyTorch to build a model that predicts product trends. This trained model is generated as a plugin.

[0153] The plugins are packaged in Docker containers and made available to field-specific organizations via an API, using a REST API that allows for easy integration with other systems.

[0154] User Rewards and Feedback

[0155] The server calculates the reward for the user who completes the quiz task. For example, 10 points are added to the user's account. The device notifies the user and displays a message saying "You've earned 10 points." The user can then check their point history in the app, which increases their motivation to participate in the next quiz task.

[0156] Specific examples

[0157] Example prompt: "Which of the following will be the best-selling products in 2022?" "Smartphones," "Tablets," "Laptops," "Smartwatches"

[0158] In this way, the system can collect high-quality data and effectively train generative AI models while increasing user engagement.

[0159] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0160] Processing Steps

[0161] Step 1: Launch the application and display the login screen

[0162] The device launches the application and displays a login screen, which displays fields for entering a username and password and a login button on the device screen.

[0163] Input: Login request from user

[0164] Output: A login screen appears

[0165] Step 2: Enter and submit user credentials

[0166] The user enters their username and password on the login screen. For example, the user enters "alice@example.com" and "password123" and clicks the Login button. The device sends this information to the server.

[0167] Input: The username and password entered by the user

[0168] Output: Authentication information sent to the server

[0169] Step 3: Validate credentials and generate tokens

[0170] The server checks the received authentication information against a database to verify the user's identity. If authentication is successful, the server generates a JSON Web Token (JWT). This token is used to maintain session information for the authenticated user.

[0171] Input: Authentication information received by the server

[0172] Output: Authentication result and generated JWT

[0173] Step 4: Login success notification and main screen display

[0174] The device saves the JWT received from the server and notifies the user that the login was successful. Specifically, it displays the message "Login successful" and transitions to the main screen.

[0175] Input: JWT sent from the server

[0176] Output: Login successful and main screen displayed

[0177] Step 5: Check usage history and generate quiz tasks

[0178] After logging in, the server checks the user's usage history. This history includes past quiz answer results and ad viewing history. If the server determines that it is time to view an ad based on the usage history, it generates a quiz-style task. For example, it could ask, "Which of the following products was the best-selling product in 2022?" with the options "smartphone," "tablet," "laptop," and "smartwatch."

[0179] Input: User usage history

[0180] Output: Generated quiz tasks

[0181] Step 6: Submitting and notifying quiz tasks

[0182] The server sends the generated quiz task to the terminal, which notifies the user of the quiz task and displays the quiz screen.

[0183] Input: Quiz task sent from the server

[0184] Output: The quiz screen shown to the user

[0185] Step 7: User responses and data submission

[0186] The user selects an answer from the options on the quiz screen and presses the send button. For example, they select "smartphone." The device then sends this answer data to the server.

[0187] Input: The user's selected answer

[0188] Output: Answer data sent to the server

[0189] Step 8: Receiving and analyzing response data

[0190] The server receives the submitted response data and temporarily stores it. It then analyzes the data using Python's pandas library and removes inappropriate data and outliers. Specifically, it performs a data cleaning process to generate clean data.

[0191] Input: Response data received by the server

[0192] Output: Cleaned data

[0193] Step 9: Add the clean data to the training dataset

[0194] The server adds the cleaned data to a training dataset, which is used to train a generative AI model.

[0195] Input: Cleaned data

[0196] Output: Updated training dataset

[0197] Step 10: Training the generative AI model

[0198] The server trains a generative AI model using the updated training dataset, for example, using TensorFlow or PyTorch to train a neural network model to create a model that performs a specific task (e.g., product trend prediction).

[0199] Input: Training dataset

[0200] Output: A trained generative AI model

[0201] Step 11: Generate and Provide the Plugin

[0202] The server generates trained generative AI models as plugins and packages them in Docker containers, and provides the plugins to organizations in specific fields via APIs, such as setting up endpoints using a REST API.

[0203] Input: A trained generative AI model

[0204] Output: Model provided as a plugin

[0205] Step 12: Calculating and notifying user rewards

[0206] The server calculates a reward for the user who has completed the quiz-style task, for example, adding 10 points to the user's account. The device notifies the user and displays a message saying "You have earned 10 points."

[0207] Input: Information about the completed quiz task

[0208] Output: Reward given to user and notification

[0209] (Application example 1)

[0210] 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."

[0211] Conventional data collection methods that rely on viewing advertisements have problems such as difficulty in data collection due to users' decreased willingness to view advertisements and the use of ad blockers. Another problem is that content recommendations based on individual users' interests are not possible, resulting in a poor user experience. To solve these problems, user-participation data collection methods and personalized content recommendation technologies are required.

[0212] 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.

[0213] In this invention, the server includes: means for inputting user authentication information and transmitting it to the server; means for the server to verify the authentication information and generate a login token if authenticated; means for providing authenticated users with a quiz task instead of viewing advertisements; means for the user to answer the quiz and transmit the answers to the server; means for the server to receive the answer data, remove inappropriate data, and add it to a training dataset; means for training an AI model based on the training dataset and generating the model as a plugin; means for providing the plugin to companies in a specific field via an API; means for rewarding users each time they complete a quiz task; means for generating and providing personalized content recommendation quizzes to authenticated users; and means for integrating and analyzing quiz answer data from multiple users to improve the performance of the AI ​​model. This enables user-participation data collection without advertisement viewing and personalized content recommendations based on individual user interests.

[0214] Creating definition statements

[0215] "User authentication information" refers to the identification information entered by a user to log into a system.

[0216] A "server" is a computing device that stores, processes, and manages data.

[0217] A "login token" is a temporary identifier that indicates an authenticated user's session and is used to authenticate subsequent communications.

[0218] A "quiz task" is a task in the form of a question provided to a user, and is a means for obtaining an answer from the user.

[0219] A "training dataset" is a collection of collected data used to train an AI model.

[0220] An "AI model" is an algorithm that analyzes data and performs a specific task.

[0221] A "plug-in" is a software module that is used to add specific functionality.

[0222] "API" stands for Application Program Interface, an interface for using functions between different software.

[0223] "Rewards" are incentives such as points or benefits that users can receive by completing quiz tasks.

[0224] A "personalized content recommendation quiz" is a quiz task that is individually provided based on the user's interests and concerns.

[0225] An "outlier" is a data point that is considered an anomaly in the collected data.

[0226] MODE FOR CARRYING OUT THE INVENTION

[0227] The present invention is a system that collects learning data for a generative AI by having users answer quiz-style tasks, and then trains an AI model based on this data. In this embodiment, the roles and specific operations of the user, terminal, and server are described.

[0228] User Authentication and Login

[0229] The user launches the smartphone app and logs in by entering their username and password on the login screen. This information is sent from the device to the server. The server references a database and verifies the authentication information. If authentication is successful, the server generates a login token and sends it to the device. The user is notified that the login was successful and is taken to the main screen.

[0230] Submitting a Quiz Task

[0231] After logging in, the server checks the user's usage history and generates a personalized quiz task. For example, it may ask a question such as "Which genre are you most interested in?" and provide the options "Movies," "Music," "Sports," and "News." The generated quiz task is sent to the terminal and notified to the user.

[0232] User responses and data submission

[0233] Users answer the quiz and press the submit button to send the answers from their device to the server. The server receives the answer data and removes inappropriate data and outliers. This clean data is added to the training dataset of the generative AI.

[0234] Data Processing and Validation

[0235] The server analyzes the collected response data and uses it as a learning dataset to train an AI model. This enables personalized content recommendations based on individual users' interests and preferences. The trained AI model is generated as a plugin and provided to companies in specific fields via API.

[0236] User Rewards and Feedback

[0237] Every time a user completes a quiz task, the server calculates the reward and adds it to the user's account. The terminal displays a notification to the user that points have been earned, encouraging them to participate in the next quiz task.

[0238] Hardware and software used

[0239] Hardware: Server, user's smartphone

[0240] Software: Flask (web application framework), SQLite (database), scikit-learn (machine learning library)

[0241] Specific examples

[0242] Consider the example of a prompt sentence in which a user answers a quiz on a smartphone app: "Which genre are you most interested in?"

[0243] Example prompt sentence:

[0244] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[0245] This allows users to enjoy content that interests them, and allows service providers to collect highly accurate data.

[0246] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0247] Program processing steps

[0248] Step 1:

[0249] The user launches the smartphone app and enters their username and password on the login screen. The entered authentication information is sent from the device to the server. The server receives this authentication information and authenticates the user by checking it against a database. If authentication is successful, the server generates a login token and sends it to the device. The device displays a notification to the user that login was successful and transitions to the main screen.

[0250] Input: Username, Password

[0251] Output: Login token, login success notification

[0252] Step 2:

[0253] The server checks the usage history of the logged-in user and generates a personalized quiz task based on that data. The content of the quiz task is, for example, a question such as "Which genre are you most interested in?" with multiple choices ("Movies," "Music," "Sports," "News"). The generated quiz task is sent to the terminal and notified to the user.

[0254] Input: User ID, usage history

[0255] Output: Quiz task

[0256] Step 3:

[0257] The user answers the quiz, selects one option, and presses the send button. This answer is sent from the device to the server.

[0258] Input: User's quiz answer

[0259] Output: Response data

[0260] Step 4:

[0261] The server receives the user's response data and performs analysis. Inappropriate data and outliers are removed, and the clean data is added to the training dataset of the generative AI.

[0262] Input: Response data

[0263] Output: Clean dataset

[0264] Step 5:

[0265] The server trains an AI model based on the clean data, which is then used to recommend personalized content. Once trained, the AI ​​model is generated as a plugin and made available to companies via API.

[0266] Input: Clean dataset

[0267] Output: Trained AI model, API plugin

[0268] Step 6:

[0269] When a user completes a quiz task, the server calculates the reward and adds it to the user's account. The device displays a notification to the user that points have been earned, encouraging them to participate in the quiz again.

[0270] Input: User ID, Quiz task completion information

[0271] Output: Reward points, notification

[0272] Example operation

[0273] Add specific examples of processing for each step.

[0274] Example of Step 1:

[0275] The user enters "alice@example.com" and "password123" into the smartphone app, the server authenticates, generates a login token, and returns it.

[0276] Example of Step 2:

[0277] The server checks the usage history of user "alice" and generates a quiz task asking "Which genre are you most interested in?" and provides the options "Movies," "Music," "Sports," and "News."

[0278] Example of Step 3:

[0279] The user selects "sports" and the answer data is sent to the server.

[0280] Example of Step 4:

[0281] The server analyzes the "Sports" responses, removes irrelevant data, and adds it to a clean dataset.

[0282] Example of Step 5:

[0283] The server uses the clean data collected to train an AI model and provides it to companies as a personalized content recommendation API.

[0284] Example of Step 6:

[0285] The server awards 10 points to user "alice" for completing the quiz task and displays a "Points Earned" notification.

[0286] Prompt Sentence Examples

[0287] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[0288] 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.

[0289] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state, by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. Specific embodiments of this system are described below.

[0290] System Overview

[0291] The system consists of the following main components:

[0292] 1. User authentication system

[0293] 2. Quiz Task Providing System

[0294] 3. Response data processing system

[0295] 4. Emotion Engine

[0296] 5. Learning Dataset Management System

[0297] 6. AI Model Training System

[0298] 7. Plugin Creation and Distribution System

[0299] 8. Reward Management System

[0300] User Authentication and Login

[0301] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0302] The terminal transmits the entered user authentication information to the server.

[0303] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0304] The device will notify the user that login was successful and transition to the main screen.

[0305] Submitting a Quiz Task

[0306] After logging in, the server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[0307] The server generates a quiz task such as "Which of the following products will be the best-selling in 2022?" and presents multiple options (e.g., smartphone, tablet, laptop, smartwatch).

[0308] The server sends the generated quiz task to the terminal.

[0309] The terminal notifies the user of the quiz task and displays a quiz screen.

[0310] User responses and data submission

[0311] The user selects one of the options displayed on the quiz screen and presses the submit button (e.g., selects smartphone).

[0312] The terminal transmits the selected answer to the server.

[0313] Data Processing and Sentiment Analysis

[0314] The server temporarily stores the received response data and starts data cleaning.

[0315] The server analyzes the received response data and removes inappropriate data and outliers.

[0316] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[0317] Learning dataset and AI model training

[0318] The server adds the clean data and emotion data to the AI's training dataset.

[0319] The server trains the AI ​​model based on the clean data and emotion data.

[0320] For example, quiz answers and emotional data collected from multiple users can be fed into an AI model to improve prediction accuracy.

[0321] Plugin generation and provision

[0322] The server generates a trained AI model as a generative AI plugin for a specific field.

[0323] The server provides the generated plugins to companies in specific fields or their own services via API.

[0324] Compensation management

[0325] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[0326] The terminal displays a notification to the user that points have been earned.

[0327] Specific examples

[0328] 1. User authentication and login details example

[0329] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[0330] The terminal sends this information to the server, which checks the information against a database.

[0331] If the authentication is successful, the server generates a token and returns it to the terminal.

[0332] The device will notify you that the login was successful and will transition to the main screen.

[0333] 2. Example of Quiz Task Submission Details

[0334] The server checks the usage history and emotional state of user "alice" and determines that it is time for her to watch an advertisement.

[0335] The emotion engine analyzes "Alice's" emotions in real time and determines, for example, that she is "relaxed."

[0336] The server generates a quiz task asking, "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[0337] The terminal displays the quiz and notifies the user.

[0338] 3. Detailed example of user responses and data submission

[0339] The user selects "smartphone" and presses the send button.

[0340] The terminal transmits this response data to the server.

[0341] The server analyzes the received response data and removes inappropriate data.

[0342] 4. Detailed examples of data processing and sentiment analysis

[0343] The server adds "smartphone" to the training dataset as clean data.

[0344] The emotion engine collects emotional data (e.g., "joy") from users' responses and adds it to the learning dataset.

[0345] 5. Detailed example of creating and providing a plugin

[0346] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[0347] The server provides this plugin to companies and their own services via API.

[0348] 6. Detailed Example of User Rewards and Feedback

[0349] The server calculates a reward of 10 points for "alice" and adds it to her account.

[0350] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[0351] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions to generate more effective learning data and improve the performance of the generation AI.

[0352] The processing flow will be explained below.

[0353] Step 1:

[0354] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0355] Step 2:

[0356] The terminal transmits the entered user authentication information to the server.

[0357] Step 3:

[0358] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0359] Step 4:

[0360] The device will notify the user that login was successful and transition to the main screen.

[0361] Step 5:

[0362] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[0363] Step 6:

[0364] The emotion engine analyzes the user's emotions and adjusts the content and difficulty of the quiz task according to their state. For example, if the user is relaxed, a more difficult quiz is presented, and if the user is stressed, an easier quiz is presented.

[0365] Step 7:

[0366] The server generates a quiz task that asks the question, "Which of the following products will be the best-selling in 2022?" and provides several options (e.g., smartphone, tablet, laptop, smartwatch).

[0367] Step 8:

[0368] The server sends the generated quiz task to the terminal.

[0369] Step 9:

[0370] The terminal notifies the user of the quiz task and displays a quiz screen.

[0371] Step 10:

[0372] The user selects one of the options displayed on the quiz screen, for example, "smartphone," and presses the send button.

[0373] Step 11:

[0374] The terminal transmits the selected answer to the server.

[0375] Step 12:

[0376] The server temporarily stores the received response data and starts cleaning the data.

[0377] Step 13:

[0378] The server analyzes the received response data and removes inappropriate data and outliers.

[0379] Step 14:

[0380] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[0381] Step 15:

[0382] The server adds the clean data and emotion data to the AI's training dataset.

[0383] Step 16:

[0384] The server trains the AI ​​model based on the clean data and emotion data.

[0385] Step 17:

[0386] The server generates a trained AI model as a generative AI plugin for a specific field.

[0387] Step 18:

[0388] The server provides the generated plugins to companies in specific fields or their own services via API.

[0389] Step 19:

[0390] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[0391] Step 20:

[0392] The terminal displays a notification to the user that points have been earned.

[0393] Example 2

[0394] 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."

[0395] Conventional ad viewing formats require users to passively watch ads, resulting in poor user experience and limited advertising effectiveness. Furthermore, because interactive tasks that take into account the user's emotional state are not provided, it is difficult to effectively collect data based on the user's interests. Furthermore, there are issues with quality control of collected data and optimizing training datasets, which can lead to the inclusion of inappropriate data.

[0396] 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 inputting user authentication information and transmitting it to a device on the network; means for the device on the network to verify the authentication information and generate a login token if authenticated; and means for providing the authenticated user with a quiz task instead of viewing an advertisement. This enables the user's advertisement viewing experience to be replaced with an active quiz task. The server also includes means for the user to answer the quiz and transmit the answer to a device on the network; and means for the device on the network to receive the answer data, clean the data, and add it to a training dataset. This improves quality control of collected data and enables the generation of an appropriate training dataset. The server also includes means for training an artificial intelligence model based on the training dataset and generating the model as a plugin, means for providing the plugin to entities in a specific field via an API, and means for granting a reward each time a user completes a quiz task. This improves the performance of the AI ​​model and increases user motivation to participate.

[0397] A "networked device" is an electronic device such as a computer, smartphone, or tablet that can communicate over the Internet or other networks.

[0398] "Authentication Information" refers to the username, password, and other authentication means entered by a user when logging into a system.

[0399] A "login token" is a unique identifier or security token that is generated when a server successfully authenticates a user and is used to maintain the user's login session.

[0400] A "quiz task" is a question-type task in which the user can participate interactively, and in which the user selects the correct answer from multiple options.

[0401] "Data cleaning" is the process of automatically removing outliers and irrelevant data from received data, a process that aims to improve data quality.

[0402] A "training dataset" is a collection of data used to train an artificial intelligence model, including clean data and sentiment data.

[0403] An "artificial intelligence model" is a machine learning algorithm that is trained on a collected dataset and is a computer program to automatically perform a specific task.

[0404] A "plugin" is a software module that adds specific functionality or services to other systems or applications via an API.

[0405] "Entity" is a general term that refers to a legal entity, company, or organization that provides business or services in a particular field.

[0406] "API" is an abbreviation for Application Programming Interface, an interface for sharing functions between different software applications.

[0407] "Rewards" refer to incentives such as points or benefits that are given to users each time they complete a quiz task.

[0408] System Overview

[0409] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. The system is implemented using devices (servers and terminals) on a network.

[0410] Hardware and software used

[0411] Hardware: Servers, devices (computers, smartphones, tablets)

[0412] Software: User authentication system, quiz task provision system, answer data processing system, emotion engine, learning dataset management system, AI model training system, reward management system

[0413] User Authentication and Login

[0414] The user starts the application using the device and the login screen is displayed. The user enters a username and password and sends them to the server. The server compares the received authentication information with the database, and if authentication is successful, generates a login token and sends it back to the device. The device displays a notification to the user that the login was successful and transitions to the main screen.

[0415] Submitting a Quiz Task

[0416] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz task based on that. The server generates quiz tasks such as "Which of the following products was the best-selling in 2022?" and presents multiple options. The server sends the generated quiz task to the device, and the device notifies the user of the quiz task and displays the quiz screen.

[0417] User responses and data submission

[0418] The user selects one answer from the options displayed on the quiz screen and presses the send button. The device then sends the selected answer to the server.

[0419] Data Processing and Sentiment Analysis

[0420] The server temporarily stores the received answer data and starts data cleaning. The server analyzes the received answer data and removes inappropriate data and outliers. The emotion engine collects emotional data (e.g., joy, surprise, irritation) from the user when answering the quiz and adds it to the training dataset.

[0421] Learning dataset and AI model training

[0422] The server adds the cleaned response data and sentiment data to the AI ​​training dataset, which the server uses to train the AI ​​model and improve its prediction accuracy.

[0423] Plugin generation and provision

[0424] The server generates the trained AI model as a generative AI plugin for a specific domain, and provides the generated plugin to entities in the specific domain via an API.

[0425] Compensation management

[0426] The server calculates the reward (points or rewards) for the user who completed the quiz task and adds it to the user's account. The terminal displays a notification to the user that points have been earned.

[0427] Specific examples

[0428] 1. Example of user authentication:

[0429] The user launches the app, enters "alice@example.com" and "password123," and presses the login button. The device sends the information to the server, which checks the information in the database. If authentication is successful, the server generates a token and sends it back to the device. The device notifies the user that "Login was successful" and returns to the main screen.

[0430] 2. Example of a quiz task:

[0431] The server checks the usage history and emotional state of user "alice," which is "relaxed." The server generates a quiz task asking, "Which of the following products was the best-selling product in 2022?" and presents the options "smartphone," "tablet," "laptop," and "smartwatch." The device displays the quiz screen and notifies the user.

[0432] 3. Example prompt:

[0433] "Which of the following products will be the best-selling in 2022?"

[0434] "Predict what kind of ad would be best based on the emotional data of the user when they answer the quiz."

[0435] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions, thereby generating more effective training data and improving the performance of the generation AI.

[0436] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0437] Step 1:

[0438] A user starts an application using a terminal and a login screen is displayed. The user enters a username and password. The terminal sends the entered authentication information (username and password) to the server. The username and password are used as input data, and that data is sent to the server.

[0439] Step 2:

[0440] The server compares the received authentication information with the database and generates a login token if authentication is successful. It searches the database for matching authentication information and obtains the result of successful authentication as output. If authentication is successful, a login token is generated and the server sends that login token to the terminal. The output is the login token if successful. As a specific example, the server compares the username "alice@example.com" with the password "password123" and generates a login token if successful.

[0441] Step 3:

[0442] Based on the login token received by the terminal from the server, the terminal displays a notification of successful login to the user and transitions to the main screen. The login token received from the server is used as input. As output, the terminal displays a "login successful" notification to the user and transitions to the main screen.

[0443] Step 4:

[0444] The server checks the usage history and emotional state based on the user ID. As input data, it compares past usage history and real-time emotional state data based on the user ID. As output, it obtains the confirmed usage history and the emotion analysis results. For example, the emotion engine analyzes the emotion of user "alice" as "relaxed."

[0445] Step 5:

[0446] The server generates quiz tasks by adjusting the difficulty and type of the quiz task based on the emotional state. The appropriate quiz task is determined based on the input data of the emotion analysis results and the user's usage history. The generated quiz task is obtained as output. As a specific example, the server generates a quiz question such as, "Which of the following products was the best-selling product in 2022?"

[0447] Step 6:

[0448] The server sends the generated quiz task to the terminal, and the terminal notifies the user of the quiz task and displays the quiz screen. As input, the terminal receives the quiz task sent from the server. As output, the quiz screen is displayed to the user. The terminal displays the quiz screen and notifies the user to "answer the quiz."

[0449] Step 7:

[0450] The user selects one of the options displayed on the quiz screen and presses the send button. The answer data selected by the user is collected as input data. The selected answer data is obtained as output. For example, the user selects "smartphone" and presses the send button.

[0451] Step 8:

[0452] The device sends the selected answer data to the server. As input, it receives the answer data selected by the user. As output, it obtains the answer data to be sent to the server. The device sends the answer data "smartphone" to the server.

[0453] Step 9:

[0454] The server temporarily stores the received response data and begins data cleaning. Inappropriate data and outliers are removed. The received response data is used as input. Cleaned, clean data is obtained as output. The server stores the "smartphone" as clean data and removes inappropriate data.

[0455] Step 10:

[0456] The emotion engine collects emotional data when users answer quizzes and adds it to the training dataset. It uses real-time emotional data when users answer quizzes as input. It obtains the collected emotional data as output. For example, it collects the user's emotion when answering a quiz as "joy."

[0457] Step 11:

[0458] The server adds the cleaned response data and emotion data to the AI's training dataset. The cleaned data and emotion data are used as input. The output is an updated training dataset.

[0459] Step 12:

[0460] The server trains the AI ​​model based on the training dataset to improve prediction accuracy. The training dataset is used as input. The output is a trained AI model.

[0461] Step 13:

[0462] The server generates a trained AI model as a generative AI plugin for a specific field. The trained AI model is used as input. The generated plugin is obtained as output.

[0463] Step 14:

[0464] The server provides the generated plugin to a domain-specific entity via an API, uses the generated plugin as input, and sends the provided plugin to the domain-specific entity as output.

[0465] Step 15:

[0466] The server calculates a reward for a user who completes a quiz task and adds it to the user's account. As input, it uses the quiz task completion data. As output, the calculated reward is added to the user's account. The server calculates 10 points for "alice" and adds it to the account.

[0467] Step 16:

[0468] The terminal displays a notification to the user that a reward has been earned. As input, it receives the reward data sent from the server. As output, it displays a notification to the user that a reward has been earned. The terminal notifies the user that "points have been earned," motivating the user to participate in the next quiz task.

[0469] (Application example 2)

[0470] 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."

[0471] Conventional advertising viewing methods have made it difficult to effectively deliver ads to users and accurately measure ad performance. Furthermore, they have been unable to optimize ads based on the user's emotional state, leaving a lack of means to improve the user's advertising experience. The present invention aims to achieve more effective and personalized ad delivery by measuring advertising effectiveness through quiz-style tasks and analyzing the user's emotional state.

[0472] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0473] In this invention, the server includes means for inputting user authentication information and transmitting it to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz task instead of watching an advertisement, means for the user to answer the quiz and transmit the answer to the server, means for the server to receive the answer data, remove inappropriate data, and add the answer to a training dataset, means for the server to analyze the user's emotional data when answering the quiz and add the data to the training dataset, means for training an AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to companies in specific fields via an API, and means for awarding a reward each time a user completes a quiz task. This makes it possible to analyze the user's emotional state and provide more effective advertisement delivery and a personalized user experience.

[0474] "User authentication information" is information for identifying users and controlling access.

[0475] A "login token" is a temporary session identifier for an authenticated user to access a system.

[0476] A "quiz task" is a challenge presented to a user consisting of a series of questions and multiple-choice options.

[0477] "Answer data" is information about the answer selected by the user for the quiz task.

[0478] "Inappropriate data" is data that is not worthy or reliable for the system to include in the training dataset.

[0479] "Emotion data" is data that indicates the user's psychological state and emotions.

[0480] A "training dataset" is a set of clean data and sentiment data used to train an AI model.

[0481] An "AI model" is a software architecture that is trained to solve a specific problem based on machine learning algorithms.

[0482] A "plug-in" is a software component that extends or adds specific functionality.

[0483] "API" stands for Application Programming Interface, an interface for exchanging data and functions between different software systems.

[0484] "Rewards" are incentives such as points or benefits that users receive by completing quiz tasks.

[0485] The present invention is a system that combines a user-generated AI with learning data provided by the AI ​​through quiz-style tasks instead of watching advertisements, and an emotion engine that recognizes the user's emotional state. To realize this system, the following steps must be performed:

[0486] First, the device collects user authentication information and sends it to the server. The server verifies the authentication information, and if authentication is successful, generates a login token and returns it to the device. Once authentication is complete, the server provides the user with a quiz-style task instead of watching an advertisement. This quiz task is customized based on the user's usage history and real-time emotional state.

[0487] When a user answers a quiz, the device sends the answer data to the server. The server analyzes the received answer data, removes inappropriate data and outliers, and adds the clean data to the training dataset. At the same time, the server uses an emotion engine to analyze the user's emotional data when answering the quiz and adds that data to the training dataset.

[0488] The server then trains a generative AI model based on this learning dataset. The resulting AI model is then generated as a generative AI plugin for a specific field. This plugin is then provided to companies in that field or to their own services via API.

[0489] Through this process, the server will reward users each time they complete a quiz task. The collected clean data and emotional data will improve the accuracy of the AI ​​model, allowing companies to provide more effective advertising and services.

[0490] The hardware and software used are as follows:

[0491] Smartphone: The device through which the user accesses and answers the quiz tasks.

[0492] Python: A programming language used to develop each component of the system.

[0493] Emotion analysis model: An AI model for analyzing a user's emotional state.

[0494] Server: Validates authentication information, provides quiz tasks, processes answer and sentiment data, trains AI models, generates and provides plugins, and grants rewards.

[0495] Examples:

[0496] For example, imagine a user using an app. The user logs in to the app using a smartphone. After successfully logging in, the server generates a quiz question: "What was the best-selling product in 2022?" and displays it on the smartphone. The user selects "smartphone" and submits the answer. The server analyzes the answer data and uses an emotion engine to collect the user's emotional data at the time of answering as "joy." Finally, the user receives a reward of 10 points.

[0497] Example prompts to input to a generative AI model:

[0498] "Answer the following question: Which of the following products was your top-selling item in 2022?"

[0499] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0500] Step 1:

[0501] The terminal obtains the user authentication information and sends it to the server.

[0502] Enter your username and password.

[0503] Output: Sends authentication information to the server.

[0504] How it works: A user launches an application on their smartphone and enters their username and password into the login screen. The device then securely transmits this information to the server using the HTTPS protocol.

[0505] Step 2:

[0506] The server verifies the authentication information and generates a login token if authenticated.

[0507] Input: The username and password submitted.

[0508] Output: Authentication result and login token.

[0509] What happens: The server checks the user information stored in the database against the authentication information sent. If authentication is successful, the server generates a temporary login token and sends it back to the device.

[0510] Step 3:

[0511] The server provides the authenticated user with a quiz task in exchange for viewing an advertisement.

[0512] Input: User ID and login token, usage history, emotional state.

[0513] Output: A customized quiz task.

[0514] How it works: The server checks the usage history and real-time emotional state of the authenticated user and generates a quiz task based on that. For example, it provides a question like "What will be the best-selling product in 2022?" with options (smartphone, tablet, laptop, smartwatch).

[0515] Step 4:

[0516] The terminal presents the generated quiz task to the user, and the user answers the quiz.

[0517] Input: A customized quiz task.

[0518] Output: The user's answer.

[0519] Specific operation: The device notifies the user of the quiz task, displays the question and options, and the user selects one and presses the answer button.

[0520] Step 5:

[0521] The terminal transmits the user's response data to the server.

[0522] Input: The user's answer.

[0523] Output: Send the response data to the server.

[0524] Specific operation: When the user selects an answer to the quiz and presses the send button, the device sends the answer data to the server.

[0525] Step 6:

[0526] The server analyzes the received response data and removes inappropriate data.

[0527] Input: User response data.

[0528] Output: Clean data.

[0529] Specific operation: The server validates the received response data, detects and removes inappropriate data and outliers, and only reliable data is added to the training dataset.

[0530] Step 7:

[0531] The server uses an emotion engine to analyze the emotion data of the user when answering the quiz.

[0532] Input: User response data.

[0533] Output: User emotion data.

[0534] Specific operation: The server uses an emotion engine to analyze the user's facial expressions and voice patterns when answering the quiz, and identifies emotions such as joy, surprise, and irritation.

[0535] Step 8:

[0536] The server adds the clean data and the emotion data to the training dataset.

[0537] Input: Clean data, sentiment data.

[0538] Output: Added to the training dataset.

[0539] Specific operation: The server stores the clean data and analyzed emotion data in a learning dataset and uses it as training data for the generative AI model.

[0540] Step 9:

[0541] The server trains a generative AI model based on the learning dataset and generates the model as a plugin.

[0542] Input: The training dataset.

[0543] Output: The trained AI model plugin.

[0544] What it does: The server runs machine learning algorithms on a training dataset to train a generative AI model for use in a specific domain, and outputs the completed model in the form of a plugin.

[0545] Step 10:

[0546] The server provides the generated AI plugins to companies in specific fields via API.

[0547] Input: A trained AI model plugin.

[0548] Output: Provided via API.

[0549] How it works: The server provides trained AI model plugins to companies in specific fields via an API interface, which then use the plugin's functions in their own systems.

[0550] Step 11:

[0551] The server awards a reward each time a user completes a quiz task.

[0552] Input: User ID.

[0553] Output: Reward points.

[0554] How it works: The server monitors the user's quiz task completion status and adds a set reward point to the user's account every time the task is completed. This reward motivates the user.

[0555] 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.

[0556] 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.

[0557] 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.

[0558] [Second embodiment]

[0559] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0560] 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.

[0561] 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).

[0562] 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.

[0563] 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.

[0564] 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).

[0565] 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.

[0566] 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.

[0567] 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.

[0568] 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.

[0569] 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.

[0570] 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."

[0571] The present invention is a system that provides quiz-style tasks in exchange for watching advertisements, and collects learning data for a generation AI by having users answer the quiz tasks, and trains an AI model based on this data. The following describes in natural language the roles and specific operations of the user, terminal, and server in an embodiment of the present invention.

[0572] 1. User authentication and login

[0573] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0574] The terminal transmits the entered user authentication information to the server.

[0575] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0576] The device will notify the user that login was successful and transition to the main screen.

[0577] 2. Submit the quiz task

[0578] After logging in, the server checks the usage history based on the user ID.

[0579] If the server determines that it is time to watch an ad, it generates a quiz-style learning task, such as "Which of the following products will be the best-selling in 2022?" with multiple choices.

[0580] The server sends the generated quiz task to the terminal.

[0581] The terminal notifies the user of the quiz task and displays a quiz screen.

[0582] 3. User responses and data submission

[0583] The user selects one of the options displayed (e.g., "smartphone," "tablet," "laptop," or "smartwatch") and presses the send button.

[0584] The terminal transmits the selected answer to the server.

[0585] The server temporarily stores the received response data and proceeds to the next process.

[0586] 4. Data processing and validation

[0587] The server analyzes the received response data and removes inappropriate data and outliers.

[0588] The remaining clean data is added to the AI ​​training dataset.

[0589] For example, answer data collected from multiple users to the question, "What was the best-selling product in 2022?" is integrated and provided to the generative AI as training data.

[0590] 5. Creating and Providing Plugins

[0591] The server trains the AI ​​model using the clean data.

[0592] A trained AI model can be plugged into a specific field (e.g., "product trend prediction").

[0593] The server provides the generated plugins to companies in specific fields or their own services via API.

[0594] 6. User Rewards and Feedback

[0595] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to their accounts.

[0596] The terminal displays a notification to the user that points have been earned.

[0597] Users can check the points and rewards they have earned, which will motivate them to take part in the next quiz task.

[0598] Specific examples

[0599] 1. User authentication and login details example

[0600] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[0601] The terminal sends this information to the server, which checks the information against a database.

[0602] If the authentication is successful, the server generates a token and returns it to the terminal.

[0603] The device will notify you that the login was successful and will transition to the main screen.

[0604] 2. Example of Quiz Task Submission Details

[0605] The server checks the usage history of user "alice" and determines that it is time to view an advertisement.

[0606] The server generates a quiz task: "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[0607] The terminal displays the quiz and notifies the user.

[0608] 3. Detailed example of user responses and data submission

[0609] The user selects "smartphone" and presses the send button.

[0610] The terminal transmits this response data to the server.

[0611] The server analyzes the received response data and removes inappropriate data.

[0612] 4. Detailed examples of data processing and validation

[0613] The server adds "smartphone" to the training dataset as clean data.

[0614] 5. Detailed example of creating and providing a plugin

[0615] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[0616] The server provides this plugin to companies and their own services via API.

[0617] 6. Detailed Example of User Rewards and Feedback

[0618] The server calculates a reward of 10 points for "alice" and adds it to her account.

[0619] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[0620] This allows users to participate in quizzes instead of watching ads, and also improves the performance of the generative AI.

[0621] The processing flow will be explained below.

[0622] Step 1:

[0623] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0624] Step 2:

[0625] The terminal transmits the entered user authentication information to the server.

[0626] Step 3:

[0627] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0628] Step 4:

[0629] The device will notify the user that login was successful and transition to the main screen.

[0630] Step 5:

[0631] The server checks the usage history based on the user ID and determines whether it is time to watch an advertisement.

[0632] Step 6:

[0633] If the server determines that an ad is required, it generates a quiz task. For example, it prepares a question such as, "Which of the following products was the best-selling in 2022?" and presents several options.

[0634] Step 7:

[0635] The server sends the generated quiz task to the terminal.

[0636] Step 8:

[0637] The terminal notifies the user of the quiz task and displays a quiz screen.

[0638] Step 9:

[0639] The user selects one of the options displayed on the quiz screen. For example, they select "Smartphone" and press the submit button.

[0640] Step 10:

[0641] The terminal transmits the selected answer to the server.

[0642] Step 11:

[0643] The server temporarily stores the received response data and starts data cleaning.

[0644] Step 12:

[0645] The server analyzes the received response data and removes inappropriate data and outliers.

[0646] Step 13:

[0647] The server adds the clean data to the AI's training dataset.

[0648] Step 14:

[0649] The server trains the AI ​​model based on the clean data.

[0650] Step 15:

[0651] The server generates a trained AI model as a generative AI plugin for a specific field.

[0652] Step 16:

[0653] The server provides this plugin to companies in specific fields or their own services via API.

[0654] Step 17:

[0655] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[0656] Step 18:

[0657] The terminal displays a notification to the user that points have been earned.

[0658] Example 1

[0659] 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."

[0660] Conventional ad-based services have the problem of being monotonous for users and low engagement. Furthermore, the revenue model based on ad viewing is unstable because it depends on the number of views. Furthermore, the accuracy of the collected data is low, making it difficult to obtain high-quality data, which is important for training generative AI models. A new system that can solve these issues and collect high-quality data while increasing user engagement is needed.

[0661] 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.

[0662] In this invention, the server includes means for inputting user authentication information and transmitting it from the terminal to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz-style task in exchange for watching an advertisement, means for the user to answer the quiz-style task and transmit the answer from the terminal to the server, means for the server to receive the answer data, analyze it to remove inappropriate data, and add the clean data to a training dataset, means for training a generative AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to organizations in a specific field via an API, and means for granting a reward each time a user completes a quiz-style task. This makes it possible to collect high-quality data and effectively train the generative AI model while increasing user engagement.

[0663] "User credentials" are the information a user enters to identify themselves and gain access, often consisting of a username and password.

[0664] A "terminal" is a device operated by a user to access the system, and includes, for example, a smartphone or a personal computer.

[0665] A "server" is a computer system that provides services over a network, and is a device that has the functions of data processing, storage, and responding to requests from clients.

[0666] A "login token" is authentication information that indicates that a user has been authenticated and is used to maintain a logged-in state for a certain period of time.

[0667] A "quiz-style task" is a quiz question that a user must answer, and is a task that includes multiple-choice questions and multiple options.

[0668] "Clean data" refers to high-quality data remaining after removing inappropriate data and outliers from the received response data.

[0669] A "training dataset" is a collection of data used to train a generative AI model, to which clean data has been added.

[0670] A "generative AI model" is a model of artificial intelligence that is trained based on collected data and algorithms used to perform specific tasks.

[0671] A "plug-in" is a software component with specific functionality that can be added to other software systems to extend their functionality.

[0672] "API" stands for Application Programming Interface, a standardized interface for communication between different software systems.

[0673] "Rewards" are points, special benefits, or other compensation that a user receives each time they complete a quiz-style task.

[0674] This invention is a system that collects high-quality data while increasing user engagement through a series of processes including user authentication, quiz-style task provision, data collection and analysis, generative AI model training, and a reward system. Below, we will explain the specific hardware and software usage methods and data processing and data calculation procedures.

[0675] User Authentication and Login

[0676] The device starts up, such as a smartphone or PC, and runs an application. The application displays a screen for entering user authentication information (username and password). The user enters this information and presses the login button. The device then sends the entered authentication information to the server using an HTTPS request.

[0677] The server checks the received authentication information against the database, and if authentication is successful, generates a JSON Web Token (JWT). This token is returned to the terminal, and a notification of successful login is displayed to the user, and the user is taken to the main screen.

[0678] Providing quiz-style tasks

[0679] After logging in, the server checks the user's usage history based on the user ID. If the usage history meets certain conditions, the server creates a quiz-style task instead of watching an advertisement. For example, the server generates a question such as "Which of the following products was the best-selling product in 2022?" along with options such as "smartphone," "tablet," "laptop," and "smartwatch."

[0680] The server sends this quiz task to the terminal, and the terminal notifies the user. The quiz screen is displayed on the terminal screen, and the user selects an answer from the displayed options.

[0681] User responses and data submission

[0682] The user selects an answer from the options and presses the submit button. The device sends the answer data to the server using an HTTPS request.

[0683] The server receives the response data and temporarily stores it. Next, it analyzes the data using Python's pandas library or similar to remove inappropriate data and outliers. The remaining clean data is added to the training dataset.

[0684] Training generative AI models

[0685] The server uses the clean data to train a generative AI model. For example, it trains a neural network model using frameworks such as TensorFlow or PyTorch to build a model that predicts product trends. This trained model is generated as a plugin.

[0686] The plugins are packaged in Docker containers and made available to field-specific organizations via an API, using a REST API that allows for easy integration with other systems.

[0687] User Rewards and Feedback

[0688] The server calculates the reward for the user who completes the quiz task. For example, 10 points are added to the user's account. The device notifies the user and displays a message saying "You've earned 10 points." The user can then check their point history in the app, which increases their motivation to participate in the next quiz task.

[0689] Specific examples

[0690] Example prompt: "Which of the following will be the best-selling products in 2022?" "Smartphones," "Tablets," "Laptops," "Smartwatches"

[0691] In this way, the system can collect high-quality data and effectively train generative AI models while increasing user engagement.

[0692] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0693] Processing Steps

[0694] Step 1: Launch the application and display the login screen

[0695] The device launches the application and displays a login screen, which displays fields for entering a username and password and a login button on the device screen.

[0696] Input: Login request from user

[0697] Output: A login screen appears

[0698] Step 2: Enter and submit user credentials

[0699] The user enters their username and password on the login screen. For example, the user enters "alice@example.com" and "password123" and clicks the Login button. The device sends this information to the server.

[0700] Input: The username and password entered by the user

[0701] Output: Authentication information sent to the server

[0702] Step 3: Validate credentials and generate tokens

[0703] The server checks the received authentication information against a database to verify the user's identity. If authentication is successful, the server generates a JSON Web Token (JWT). This token is used to maintain session information for the authenticated user.

[0704] Input: Authentication information received by the server

[0705] Output: Authentication result and generated JWT

[0706] Step 4: Login success notification and main screen display

[0707] The device saves the JWT received from the server and notifies the user that the login was successful. Specifically, it displays the message "Login successful" and transitions to the main screen.

[0708] Input: JWT sent from the server

[0709] Output: Login successful and main screen displayed

[0710] Step 5: Check usage history and generate quiz tasks

[0711] After logging in, the server checks the user's usage history. This history includes past quiz answer results and ad viewing history. If the server determines that it is time to view an ad based on the usage history, it generates a quiz-style task. For example, it could ask, "Which of the following products was the best-selling product in 2022?" with the options "smartphone," "tablet," "laptop," and "smartwatch."

[0712] Input: User usage history

[0713] Output: Generated quiz tasks

[0714] Step 6: Submitting and notifying quiz tasks

[0715] The server sends the generated quiz task to the terminal, which notifies the user of the quiz task and displays the quiz screen.

[0716] Input: Quiz task sent from the server

[0717] Output: The quiz screen shown to the user

[0718] Step 7: User responses and data submission

[0719] The user selects an answer from the options on the quiz screen and presses the send button. For example, they select "smartphone." The device then sends this answer data to the server.

[0720] Input: The user's selected answer

[0721] Output: Answer data sent to the server

[0722] Step 8: Receiving and analyzing response data

[0723] The server receives the submitted response data and temporarily stores it. It then analyzes the data using Python's pandas library and removes inappropriate data and outliers. Specifically, it performs a data cleaning process to generate clean data.

[0724] Input: Response data received by the server

[0725] Output: Cleaned data

[0726] Step 9: Add the clean data to the training dataset

[0727] The server adds the cleaned data to a training dataset, which is used to train a generative AI model.

[0728] Input: Cleaned data

[0729] Output: Updated training dataset

[0730] Step 10: Training the generative AI model

[0731] The server trains a generative AI model using the updated training dataset, for example, using TensorFlow or PyTorch to train a neural network model to create a model that performs a specific task (e.g., product trend prediction).

[0732] Input: Training dataset

[0733] Output: A trained generative AI model

[0734] Step 11: Generate and Provide the Plugin

[0735] The server generates trained generative AI models as plugins and packages them in Docker containers, and provides the plugins to organizations in specific fields via APIs, such as setting up endpoints using a REST API.

[0736] Input: A trained generative AI model

[0737] Output: Model provided as a plugin

[0738] Step 12: Calculating and notifying user rewards

[0739] The server calculates a reward for the user who has completed the quiz-style task, for example, adding 10 points to the user's account. The device notifies the user and displays a message saying "You have earned 10 points."

[0740] Input: Information about the completed quiz task

[0741] Output: Reward given to user and notification

[0742] (Application example 1)

[0743] 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."

[0744] Conventional data collection methods that rely on viewing advertisements have problems such as difficulty in data collection due to users' decreased willingness to view advertisements and the use of ad blockers. Another problem is that content recommendations based on individual users' interests are not possible, resulting in a poor user experience. To solve these problems, user-participation data collection methods and personalized content recommendation technologies are required.

[0745] 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.

[0746] In this invention, the server includes: means for inputting user authentication information and transmitting it to the server; means for the server to verify the authentication information and generate a login token if authenticated; means for providing authenticated users with a quiz task instead of viewing advertisements; means for the user to answer the quiz and transmit the answers to the server; means for the server to receive the answer data, remove inappropriate data, and add it to a training dataset; means for training an AI model based on the training dataset and generating the model as a plugin; means for providing the plugin to companies in a specific field via an API; means for rewarding users each time they complete a quiz task; means for generating and providing personalized content recommendation quizzes to authenticated users; and means for integrating and analyzing quiz answer data from multiple users to improve the performance of the AI ​​model. This enables user-participation data collection without advertisement viewing and personalized content recommendations based on individual user interests.

[0747] Creating definition statements

[0748] "User authentication information" refers to the identification information entered by a user to log into a system.

[0749] A "server" is a computing device that stores, processes, and manages data.

[0750] A "login token" is a temporary identifier that indicates an authenticated user's session and is used to authenticate subsequent communications.

[0751] A "quiz task" is a task in the form of a question provided to a user, and is a means for obtaining an answer from the user.

[0752] A "training dataset" is a collection of collected data used to train an AI model.

[0753] An "AI model" is an algorithm that analyzes data and performs a specific task.

[0754] A "plug-in" is a software module that is used to add specific functionality.

[0755] "API" stands for Application Program Interface, an interface for using functions between different software.

[0756] "Rewards" are incentives such as points or benefits that users can receive by completing quiz tasks.

[0757] A "personalized content recommendation quiz" is a quiz task that is individually provided based on the user's interests and concerns.

[0758] An "outlier" is a data point that is considered an anomaly in the collected data.

[0759] MODE FOR CARRYING OUT THE INVENTION

[0760] The present invention is a system that collects learning data for a generative AI by having users answer quiz-style tasks, and then trains an AI model based on this data. In this embodiment, the roles and specific operations of the user, terminal, and server are described.

[0761] User Authentication and Login

[0762] The user launches the smartphone app and logs in by entering their username and password on the login screen. This information is sent from the device to the server. The server references a database and verifies the authentication information. If authentication is successful, the server generates a login token and sends it to the device. The user is notified that the login was successful and is taken to the main screen.

[0763] Submitting a Quiz Task

[0764] After logging in, the server checks the user's usage history and generates a personalized quiz task. For example, it may ask a question such as "Which genre are you most interested in?" and provide the options "Movies," "Music," "Sports," and "News." The generated quiz task is sent to the terminal and notified to the user.

[0765] User responses and data submission

[0766] Users answer the quiz and press the submit button to send the answers from their device to the server. The server receives the answer data and removes inappropriate data and outliers. This clean data is added to the training dataset of the generative AI.

[0767] Data Processing and Validation

[0768] The server analyzes the collected response data and uses it as a learning dataset to train an AI model. This enables personalized content recommendations based on individual users' interests and preferences. The trained AI model is generated as a plugin and provided to companies in specific fields via API.

[0769] User Rewards and Feedback

[0770] Every time a user completes a quiz task, the server calculates the reward and adds it to the user's account. The terminal displays a notification to the user that points have been earned, encouraging them to participate in the next quiz task.

[0771] Hardware and software used

[0772] Hardware: Server, user's smartphone

[0773] Software: Flask (web application framework), SQLite (database), scikit-learn (machine learning library)

[0774] Specific examples

[0775] Consider the example of a prompt sentence in which a user answers a quiz on a smartphone app: "Which genre are you most interested in?"

[0776] Example prompt sentence:

[0777] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[0778] This allows users to enjoy content that interests them, and allows service providers to collect highly accurate data.

[0779] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0780] Program processing steps

[0781] Step 1:

[0782] The user launches the smartphone app and enters their username and password on the login screen. The entered authentication information is sent from the device to the server. The server receives this authentication information and authenticates the user by checking it against a database. If authentication is successful, the server generates a login token and sends it to the device. The device displays a notification to the user that login was successful and transitions to the main screen.

[0783] Input: Username, Password

[0784] Output: Login token, login success notification

[0785] Step 2:

[0786] The server checks the usage history of the logged-in user and generates a personalized quiz task based on that data. The content of the quiz task is, for example, a question such as "Which genre are you most interested in?" with multiple choices ("Movies," "Music," "Sports," "News"). The generated quiz task is sent to the terminal and notified to the user.

[0787] Input: User ID, usage history

[0788] Output: Quiz task

[0789] Step 3:

[0790] The user answers the quiz, selects one option, and presses the send button. This answer is sent from the device to the server.

[0791] Input: User's quiz answer

[0792] Output: Response data

[0793] Step 4:

[0794] The server receives the user's response data and performs analysis. Inappropriate data and outliers are removed, and the clean data is added to the training dataset of the generative AI.

[0795] Input: Response data

[0796] Output: Clean dataset

[0797] Step 5:

[0798] The server trains an AI model based on the clean data, which is then used to recommend personalized content. Once trained, the AI ​​model is generated as a plugin and made available to companies via API.

[0799] Input: Clean dataset

[0800] Output: Trained AI model, API plugin

[0801] Step 6:

[0802] When a user completes a quiz task, the server calculates the reward and adds it to the user's account. The device displays a notification to the user that points have been earned, encouraging them to participate in the quiz again.

[0803] Input: User ID, Quiz task completion information

[0804] Output: Reward points, notification

[0805] Example operation

[0806] Add specific examples of processing for each step.

[0807] Example of Step 1:

[0808] The user enters "alice@example.com" and "password123" into the smartphone app, the server authenticates, generates a login token, and returns it.

[0809] Example of Step 2:

[0810] The server checks the usage history of user "alice" and generates a quiz task asking "Which genre are you most interested in?" and provides the options "Movies," "Music," "Sports," and "News."

[0811] Example of Step 3:

[0812] The user selects "sports" and the answer data is sent to the server.

[0813] Example of Step 4:

[0814] The server analyzes the "Sports" responses, removes irrelevant data, and adds it to a clean dataset.

[0815] Example of Step 5:

[0816] The server uses the clean data collected to train an AI model and provides it to companies as a personalized content recommendation API.

[0817] Example of Step 6:

[0818] The server awards 10 points to user "alice" for completing the quiz task and displays a "Points Earned" notification.

[0819] Prompt Sentence Examples

[0820] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[0821] 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.

[0822] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state, by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. Specific embodiments of this system are described below.

[0823] System Overview

[0824] The system consists of the following main components:

[0825] 1. User authentication system

[0826] 2. Quiz Task Providing System

[0827] 3. Response data processing system

[0828] 4. Emotion Engine

[0829] 5. Learning Dataset Management System

[0830] 6. AI Model Training System

[0831] 7. Plugin Creation and Distribution System

[0832] 8. Reward Management System

[0833] User Authentication and Login

[0834] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0835] The terminal transmits the entered user authentication information to the server.

[0836] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0837] The device will notify the user that login was successful and transition to the main screen.

[0838] Submitting a Quiz Task

[0839] After logging in, the server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[0840] The server generates a quiz task such as "Which of the following products will be the best-selling in 2022?" and presents multiple options (e.g., smartphone, tablet, laptop, smartwatch).

[0841] The server sends the generated quiz task to the terminal.

[0842] The terminal notifies the user of the quiz task and displays a quiz screen.

[0843] User responses and data submission

[0844] The user selects one of the options displayed on the quiz screen and presses the submit button (e.g., selects smartphone).

[0845] The terminal transmits the selected answer to the server.

[0846] Data Processing and Sentiment Analysis

[0847] The server temporarily stores the received response data and starts data cleaning.

[0848] The server analyzes the received response data and removes inappropriate data and outliers.

[0849] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[0850] Learning dataset and AI model training

[0851] The server adds the clean data and emotion data to the AI's training dataset.

[0852] The server trains the AI ​​model based on the clean data and emotion data.

[0853] For example, quiz answers and emotional data collected from multiple users can be fed into an AI model to improve prediction accuracy.

[0854] Plugin generation and provision

[0855] The server generates a trained AI model as a generative AI plugin for a specific field.

[0856] The server provides the generated plugins to companies in specific fields or their own services via API.

[0857] Compensation management

[0858] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[0859] The terminal displays a notification to the user that points have been earned.

[0860] Specific examples

[0861] 1. User authentication and login details example

[0862] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[0863] The terminal sends this information to the server, which checks the information against a database.

[0864] If the authentication is successful, the server generates a token and returns it to the terminal.

[0865] The device will notify you that the login was successful and will transition to the main screen.

[0866] 2. Example of Quiz Task Submission Details

[0867] The server checks the usage history and emotional state of user "alice" and determines that it is time for her to watch an advertisement.

[0868] The emotion engine analyzes "Alice's" emotions in real time and determines, for example, that she is "relaxed."

[0869] The server generates a quiz task asking, "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[0870] The terminal displays the quiz and notifies the user.

[0871] 3. Detailed example of user responses and data submission

[0872] The user selects "smartphone" and presses the send button.

[0873] The terminal transmits this response data to the server.

[0874] The server analyzes the received response data and removes inappropriate data.

[0875] 4. Detailed examples of data processing and sentiment analysis

[0876] The server adds "smartphone" to the training dataset as clean data.

[0877] The emotion engine collects emotional data (e.g., "joy") from users' responses and adds it to the learning dataset.

[0878] 5. Detailed example of creating and providing a plugin

[0879] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[0880] The server provides this plugin to companies and their own services via API.

[0881] 6. Detailed Example of User Rewards and Feedback

[0882] The server calculates a reward of 10 points for "alice" and adds it to her account.

[0883] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[0884] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions to generate more effective learning data and improve the performance of the generation AI.

[0885] The processing flow will be explained below.

[0886] Step 1:

[0887] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[0888] Step 2:

[0889] The terminal transmits the entered user authentication information to the server.

[0890] Step 3:

[0891] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[0892] Step 4:

[0893] The device will notify the user that login was successful and transition to the main screen.

[0894] Step 5:

[0895] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[0896] Step 6:

[0897] The emotion engine analyzes the user's emotions and adjusts the content and difficulty of the quiz task according to their state. For example, if the user is relaxed, a more difficult quiz is presented, and if the user is stressed, an easier quiz is presented.

[0898] Step 7:

[0899] The server generates a quiz task that asks the question, "Which of the following products will be the best-selling in 2022?" and provides several options (e.g., smartphone, tablet, laptop, smartwatch).

[0900] Step 8:

[0901] The server sends the generated quiz task to the terminal.

[0902] Step 9:

[0903] The terminal notifies the user of the quiz task and displays a quiz screen.

[0904] Step 10:

[0905] The user selects one of the options displayed on the quiz screen, for example, "smartphone," and presses the send button.

[0906] Step 11:

[0907] The terminal transmits the selected answer to the server.

[0908] Step 12:

[0909] The server temporarily stores the received response data and starts cleaning the data.

[0910] Step 13:

[0911] The server analyzes the received response data and removes inappropriate data and outliers.

[0912] Step 14:

[0913] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[0914] Step 15:

[0915] The server adds the clean data and emotion data to the AI's training dataset.

[0916] Step 16:

[0917] The server trains the AI ​​model based on the clean data and emotion data.

[0918] Step 17:

[0919] The server generates a trained AI model as a generative AI plugin for a specific field.

[0920] Step 18:

[0921] The server provides the generated plugins to companies in specific fields or their own services via API.

[0922] Step 19:

[0923] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[0924] Step 20:

[0925] The terminal displays a notification to the user that points have been earned.

[0926] Example 2

[0927] 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."

[0928] Conventional ad viewing formats require users to passively watch ads, resulting in poor user experience and limited advertising effectiveness. Furthermore, because interactive tasks that take into account the user's emotional state are not provided, it is difficult to effectively collect data based on the user's interests. Furthermore, there are issues with quality control of collected data and optimizing training datasets, which can lead to the inclusion of inappropriate data.

[0929] 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 inputting user authentication information and transmitting it to a device on the network; means for the device on the network to verify the authentication information and generate a login token if authenticated; and means for providing the authenticated user with a quiz task instead of viewing an advertisement. This enables the user's advertisement viewing experience to be replaced with an active quiz task. The server also includes means for the user to answer the quiz and transmit the answer to a device on the network; and means for the device on the network to receive the answer data, clean the data, and add it to a training dataset. This improves quality control of collected data and enables the generation of an appropriate training dataset. The server also includes means for training an artificial intelligence model based on the training dataset and generating the model as a plugin, means for providing the plugin to entities in a specific field via an API, and means for granting a reward each time a user completes a quiz task. This improves the performance of the AI ​​model and increases user motivation to participate.

[0930] A "networked device" is an electronic device such as a computer, smartphone, or tablet that can communicate over the Internet or other networks.

[0931] "Authentication Information" refers to the username, password, and other authentication means entered by a user when logging into a system.

[0932] A "login token" is a unique identifier or security token that is generated when a server successfully authenticates a user and is used to maintain the user's login session.

[0933] A "quiz task" is a question-type task in which the user can participate interactively, and in which the user selects the correct answer from multiple options.

[0934] "Data cleaning" is the process of automatically removing outliers and irrelevant data from received data, a process that aims to improve data quality.

[0935] A "training dataset" is a collection of data used to train an artificial intelligence model, including clean data and sentiment data.

[0936] An "artificial intelligence model" is a machine learning algorithm that is trained on a collected dataset and is a computer program to automatically perform a specific task.

[0937] A "plugin" is a software module that adds specific functionality or services to other systems or applications via an API.

[0938] "Entity" is a general term that refers to a legal entity, company, or organization that provides business or services in a particular field.

[0939] "API" is an abbreviation for Application Programming Interface, an interface for sharing functions between different software applications.

[0940] "Rewards" refer to incentives such as points or benefits that are given to users each time they complete a quiz task.

[0941] System Overview

[0942] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. The system is implemented using devices (servers and terminals) on a network.

[0943] Hardware and software used

[0944] Hardware: Servers, devices (computers, smartphones, tablets)

[0945] Software: User authentication system, quiz task provision system, answer data processing system, emotion engine, learning dataset management system, AI model training system, reward management system

[0946] User Authentication and Login

[0947] The user starts the application using the device and the login screen is displayed. The user enters a username and password and sends them to the server. The server compares the received authentication information with the database, and if authentication is successful, generates a login token and sends it back to the device. The device displays a notification to the user that the login was successful and transitions to the main screen.

[0948] Submitting a Quiz Task

[0949] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz task based on that. The server generates quiz tasks such as "Which of the following products was the best-selling in 2022?" and presents multiple options. The server sends the generated quiz task to the device, and the device notifies the user of the quiz task and displays the quiz screen.

[0950] User responses and data submission

[0951] The user selects one answer from the options displayed on the quiz screen and presses the send button. The device then sends the selected answer to the server.

[0952] Data Processing and Sentiment Analysis

[0953] The server temporarily stores the received answer data and starts data cleaning. The server analyzes the received answer data and removes inappropriate data and outliers. The emotion engine collects emotional data (e.g., joy, surprise, irritation) from the user when answering the quiz and adds it to the training dataset.

[0954] Learning dataset and AI model training

[0955] The server adds the cleaned response data and sentiment data to the AI ​​training dataset, which the server uses to train the AI ​​model and improve its prediction accuracy.

[0956] Plugin generation and provision

[0957] The server generates the trained AI model as a generative AI plugin for a specific domain, and provides the generated plugin to entities in the specific domain via an API.

[0958] Compensation management

[0959] The server calculates the reward (points or rewards) for the user who completed the quiz task and adds it to the user's account. The terminal displays a notification to the user that points have been earned.

[0960] Specific examples

[0961] 1. Example of user authentication:

[0962] The user launches the app, enters "alice@example.com" and "password123," and presses the login button. The device sends the information to the server, which checks the information in the database. If authentication is successful, the server generates a token and sends it back to the device. The device notifies the user that "Login was successful" and returns to the main screen.

[0963] 2. Example of a quiz task:

[0964] The server checks the usage history and emotional state of user "alice," which is "relaxed." The server generates a quiz task asking, "Which of the following products was the best-selling product in 2022?" and presents the options "smartphone," "tablet," "laptop," and "smartwatch." The device displays the quiz screen and notifies the user.

[0965] 3. Example prompt:

[0966] "Which of the following products will be the best-selling in 2022?"

[0967] "Predict what kind of ad would be best based on the emotional data of the user when they answer the quiz."

[0968] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions, thereby generating more effective training data and improving the performance of the generation AI.

[0969] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0970] Step 1:

[0971] A user starts an application using a terminal and a login screen is displayed. The user enters a username and password. The terminal sends the entered authentication information (username and password) to the server. The username and password are used as input data, and that data is sent to the server.

[0972] Step 2:

[0973] The server compares the received authentication information with the database and generates a login token if authentication is successful. It searches the database for matching authentication information and obtains the result of successful authentication as output. If authentication is successful, a login token is generated and the server sends that login token to the terminal. The output is the login token if successful. As a specific example, the server compares the username "alice@example.com" with the password "password123" and generates a login token if successful.

[0974] Step 3:

[0975] Based on the login token received by the terminal from the server, the terminal displays a notification of successful login to the user and transitions to the main screen. The login token received from the server is used as input. As output, the terminal displays a "login successful" notification to the user and transitions to the main screen.

[0976] Step 4:

[0977] The server checks the usage history and emotional state based on the user ID. As input data, it compares past usage history and real-time emotional state data based on the user ID. As output, it obtains the confirmed usage history and the emotion analysis results. For example, the emotion engine analyzes the emotion of user "alice" as "relaxed."

[0978] Step 5:

[0979] The server generates quiz tasks by adjusting the difficulty and type of the quiz task based on the emotional state. The appropriate quiz task is determined based on the input data of the emotion analysis results and the user's usage history. The generated quiz task is obtained as output. As a specific example, the server generates a quiz question such as, "Which of the following products was the best-selling product in 2022?"

[0980] Step 6:

[0981] The server sends the generated quiz task to the terminal, and the terminal notifies the user of the quiz task and displays the quiz screen. As input, the terminal receives the quiz task sent from the server. As output, the quiz screen is displayed to the user. The terminal displays the quiz screen and notifies the user to "answer the quiz."

[0982] Step 7:

[0983] The user selects one of the options displayed on the quiz screen and presses the send button. The answer data selected by the user is collected as input data. The selected answer data is obtained as output. For example, the user selects "smartphone" and presses the send button.

[0984] Step 8:

[0985] The device sends the selected answer data to the server. As input, it receives the answer data selected by the user. As output, it obtains the answer data to be sent to the server. The device sends the answer data "smartphone" to the server.

[0986] Step 9:

[0987] The server temporarily stores the received response data and begins data cleaning. Inappropriate data and outliers are removed. The received response data is used as input. Cleaned, clean data is obtained as output. The server stores the "smartphone" as clean data and removes inappropriate data.

[0988] Step 10:

[0989] The emotion engine collects emotional data when users answer quizzes and adds it to the training dataset. It uses real-time emotional data when users answer quizzes as input. It obtains the collected emotional data as output. For example, it collects the user's emotion when answering a quiz as "joy."

[0990] Step 11:

[0991] The server adds the cleaned response data and emotion data to the AI's training dataset. The cleaned data and emotion data are used as input. The output is an updated training dataset.

[0992] Step 12:

[0993] The server trains the AI ​​model based on the training dataset to improve prediction accuracy. The training dataset is used as input. The output is a trained AI model.

[0994] Step 13:

[0995] The server generates a trained AI model as a generative AI plugin for a specific field. The trained AI model is used as input. The generated plugin is obtained as output.

[0996] Step 14:

[0997] The server provides the generated plugin to a domain-specific entity via an API, uses the generated plugin as input, and sends the provided plugin to the domain-specific entity as output.

[0998] Step 15:

[0999] The server calculates a reward for a user who completes a quiz task and adds it to the user's account. As input, it uses the quiz task completion data. As output, the calculated reward is added to the user's account. The server calculates 10 points for "alice" and adds it to the account.

[1000] Step 16:

[1001] The terminal displays a notification to the user that a reward has been earned. As input, it receives the reward data sent from the server. As output, it displays a notification to the user that a reward has been earned. The terminal notifies the user that "points have been earned," motivating the user to participate in the next quiz task.

[1002] (Application example 2)

[1003] 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."

[1004] Conventional advertising viewing methods have made it difficult to effectively deliver ads to users and accurately measure ad performance. Furthermore, they have been unable to optimize ads based on the user's emotional state, leaving a lack of means to improve the user's advertising experience. The present invention aims to achieve more effective and personalized ad delivery by measuring advertising effectiveness through quiz-style tasks and analyzing the user's emotional state.

[1005] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1006] In this invention, the server includes means for inputting user authentication information and transmitting it to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz task instead of watching an advertisement, means for the user to answer the quiz and transmit the answer to the server, means for the server to receive the answer data, remove inappropriate data, and add the answer to a training dataset, means for the server to analyze the user's emotional data when answering the quiz and add the data to the training dataset, means for training an AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to companies in specific fields via an API, and means for awarding a reward each time a user completes a quiz task. This makes it possible to analyze the user's emotional state and provide more effective advertisement delivery and a personalized user experience.

[1007] "User authentication information" is information for identifying users and controlling access.

[1008] A "login token" is a temporary session identifier for an authenticated user to access a system.

[1009] A "quiz task" is a challenge presented to a user consisting of a series of questions and multiple-choice options.

[1010] "Answer data" is information about the answer selected by the user for the quiz task.

[1011] "Inappropriate data" is data that is not worthy or reliable for the system to include in the training dataset.

[1012] "Emotion data" is data that indicates the user's psychological state and emotions.

[1013] A "training dataset" is a set of clean data and sentiment data used to train an AI model.

[1014] An "AI model" is a software architecture that is trained to solve a specific problem based on machine learning algorithms.

[1015] A "plug-in" is a software component that extends or adds specific functionality.

[1016] "API" stands for Application Programming Interface, an interface for exchanging data and functions between different software systems.

[1017] "Rewards" are incentives such as points or benefits that users receive by completing quiz tasks.

[1018] The present invention is a system that combines a user-generated AI with learning data provided by the AI ​​through quiz-style tasks instead of watching advertisements, and an emotion engine that recognizes the user's emotional state. To realize this system, the following steps must be performed:

[1019] First, the device collects user authentication information and sends it to the server. The server verifies the authentication information, and if authentication is successful, generates a login token and returns it to the device. Once authentication is complete, the server provides the user with a quiz-style task instead of watching an advertisement. This quiz task is customized based on the user's usage history and real-time emotional state.

[1020] When a user answers a quiz, the device sends the answer data to the server. The server analyzes the received answer data, removes inappropriate data and outliers, and adds the clean data to the training dataset. At the same time, the server uses an emotion engine to analyze the user's emotional data when answering the quiz and adds that data to the training dataset.

[1021] The server then trains a generative AI model based on this learning dataset. The resulting AI model is then generated as a generative AI plugin for a specific field. This plugin is then provided to companies in that field or to their own services via API.

[1022] Through this process, the server will reward users each time they complete a quiz task. The collected clean data and emotional data will improve the accuracy of the AI ​​model, allowing companies to provide more effective advertising and services.

[1023] The hardware and software used are as follows:

[1024] Smartphone: The device through which the user accesses and answers the quiz tasks.

[1025] Python: A programming language used to develop each component of the system.

[1026] Emotion analysis model: An AI model for analyzing a user's emotional state.

[1027] Server: Validates authentication information, provides quiz tasks, processes answer and sentiment data, trains AI models, generates and provides plugins, and grants rewards.

[1028] Examples:

[1029] For example, imagine a user using an app. The user logs in to the app using a smartphone. After successfully logging in, the server generates a quiz question: "What was the best-selling product in 2022?" and displays it on the smartphone. The user selects "smartphone" and submits the answer. The server analyzes the answer data and uses an emotion engine to collect the user's emotional data at the time of answering as "joy." Finally, the user receives a reward of 10 points.

[1030] Example prompts to input to a generative AI model:

[1031] "Answer the following question: Which of the following products was your top-selling item in 2022?"

[1032] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1033] Step 1:

[1034] The terminal obtains the user authentication information and sends it to the server.

[1035] Enter your username and password.

[1036] Output: Sends authentication information to the server.

[1037] How it works: A user launches an application on their smartphone and enters their username and password into the login screen. The device then securely transmits this information to the server using the HTTPS protocol.

[1038] Step 2:

[1039] The server verifies the authentication information and generates a login token if authenticated.

[1040] Input: The username and password submitted.

[1041] Output: Authentication result and login token.

[1042] What happens: The server checks the user information stored in the database against the authentication information sent. If authentication is successful, the server generates a temporary login token and sends it back to the device.

[1043] Step 3:

[1044] The server provides the authenticated user with a quiz task in exchange for viewing an advertisement.

[1045] Input: User ID and login token, usage history, emotional state.

[1046] Output: A customized quiz task.

[1047] How it works: The server checks the usage history and real-time emotional state of the authenticated user and generates a quiz task based on that. For example, it provides a question like "What will be the best-selling product in 2022?" with options (smartphone, tablet, laptop, smartwatch).

[1048] Step 4:

[1049] The terminal presents the generated quiz task to the user, and the user answers the quiz.

[1050] Input: A customized quiz task.

[1051] Output: The user's answer.

[1052] Specific operation: The device notifies the user of the quiz task, displays the question and options, and the user selects one and presses the answer button.

[1053] Step 5:

[1054] The terminal transmits the user's response data to the server.

[1055] Input: The user's answer.

[1056] Output: Send the response data to the server.

[1057] Specific operation: When the user selects an answer to the quiz and presses the send button, the device sends the answer data to the server.

[1058] Step 6:

[1059] The server analyzes the received response data and removes inappropriate data.

[1060] Input: User response data.

[1061] Output: Clean data.

[1062] Specific operation: The server validates the received response data, detects and removes inappropriate data and outliers, and only reliable data is added to the training dataset.

[1063] Step 7:

[1064] The server uses an emotion engine to analyze the emotion data of the user when answering the quiz.

[1065] Input: User response data.

[1066] Output: User emotion data.

[1067] Specific operation: The server uses an emotion engine to analyze the user's facial expressions and voice patterns when answering the quiz, and identifies emotions such as joy, surprise, and irritation.

[1068] Step 8:

[1069] The server adds the clean data and the emotion data to the training dataset.

[1070] Input: Clean data, sentiment data.

[1071] Output: Added to the training dataset.

[1072] Specific operation: The server stores the clean data and analyzed emotion data in a learning dataset and uses it as training data for the generative AI model.

[1073] Step 9:

[1074] The server trains a generative AI model based on the learning dataset and generates the model as a plugin.

[1075] Input: The training dataset.

[1076] Output: The trained AI model plugin.

[1077] What it does: The server runs machine learning algorithms on a training dataset to train a generative AI model for use in a specific domain, and outputs the completed model in the form of a plugin.

[1078] Step 10:

[1079] The server provides the generated AI plugins to companies in specific fields via API.

[1080] Input: A trained AI model plugin.

[1081] Output: Provided via API.

[1082] How it works: The server provides trained AI model plugins to companies in specific fields via an API interface, which then use the plugin's functions in their own systems.

[1083] Step 11:

[1084] The server awards a reward each time a user completes a quiz task.

[1085] Input: User ID.

[1086] Output: Reward points.

[1087] How it works: The server monitors the user's quiz task completion status and adds a set reward point to the user's account every time the task is completed. This reward motivates the user.

[1088] 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.

[1089] 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.

[1090] 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.

[1091] [Third embodiment]

[1092] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1093] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1094] 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).

[1095] 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.

[1096] 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.

[1097] 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).

[1098] 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.

[1099] 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.

[1100] 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.

[1101] 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.

[1102] 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.

[1103] 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."

[1104] The present invention is a system that provides quiz-style tasks in exchange for watching advertisements, and collects learning data for a generation AI by having users answer the quiz tasks, and trains an AI model based on this data. The following describes in natural language the roles and specific operations of the user, terminal, and server in an embodiment of the present invention.

[1105] 1. User authentication and login

[1106] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1107] The terminal transmits the entered user authentication information to the server.

[1108] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1109] The device will notify the user that login was successful and transition to the main screen.

[1110] 2. Submit the quiz task

[1111] After logging in, the server checks the usage history based on the user ID.

[1112] If the server determines that it is time to watch an ad, it generates a quiz-style learning task, such as "Which of the following products will be the best-selling in 2022?" with multiple choices.

[1113] The server sends the generated quiz task to the terminal.

[1114] The terminal notifies the user of the quiz task and displays a quiz screen.

[1115] 3. User responses and data submission

[1116] The user selects one of the options displayed (e.g., "smartphone," "tablet," "laptop," or "smartwatch") and presses the send button.

[1117] The terminal transmits the selected answer to the server.

[1118] The server temporarily stores the received response data and proceeds to the next process.

[1119] 4. Data processing and validation

[1120] The server analyzes the received response data and removes inappropriate data and outliers.

[1121] The remaining clean data is added to the AI ​​training dataset.

[1122] For example, answer data collected from multiple users to the question, "What was the best-selling product in 2022?" is integrated and provided to the generative AI as training data.

[1123] 5. Creating and Providing Plugins

[1124] The server trains the AI ​​model using the clean data.

[1125] A trained AI model can be plugged into a specific field (e.g., "product trend prediction").

[1126] The server provides the generated plugins to companies in specific fields or their own services via API.

[1127] 6. User Rewards and Feedback

[1128] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to their accounts.

[1129] The terminal displays a notification to the user that points have been earned.

[1130] Users can check the points and rewards they have earned, which will motivate them to take part in the next quiz task.

[1131] Specific examples

[1132] 1. User authentication and login details example

[1133] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[1134] The terminal sends this information to the server, which checks the information against a database.

[1135] If the authentication is successful, the server generates a token and returns it to the terminal.

[1136] The device will notify you that the login was successful and will transition to the main screen.

[1137] 2. Example of Quiz Task Submission Details

[1138] The server checks the usage history of user "alice" and determines that it is time to view an advertisement.

[1139] The server generates a quiz task: "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[1140] The terminal displays the quiz and notifies the user.

[1141] 3. Detailed example of user responses and data submission

[1142] The user selects "smartphone" and presses the send button.

[1143] The terminal transmits this response data to the server.

[1144] The server analyzes the received response data and removes inappropriate data.

[1145] 4. Detailed examples of data processing and validation

[1146] The server adds "smartphone" to the training dataset as clean data.

[1147] 5. Detailed example of creating and providing a plugin

[1148] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[1149] The server provides this plugin to companies and their own services via API.

[1150] 6. Detailed Example of User Rewards and Feedback

[1151] The server calculates a reward of 10 points for "alice" and adds it to her account.

[1152] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[1153] This allows users to participate in quizzes instead of watching ads, and also improves the performance of the generative AI.

[1154] The processing flow will be explained below.

[1155] Step 1:

[1156] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1157] Step 2:

[1158] The terminal transmits the entered user authentication information to the server.

[1159] Step 3:

[1160] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1161] Step 4:

[1162] The device will notify the user that login was successful and transition to the main screen.

[1163] Step 5:

[1164] The server checks the usage history based on the user ID and determines whether it is time to watch an advertisement.

[1165] Step 6:

[1166] If the server determines that an ad is required, it generates a quiz task. For example, it prepares a question such as, "Which of the following products was the best-selling in 2022?" and presents several options.

[1167] Step 7:

[1168] The server sends the generated quiz task to the terminal.

[1169] Step 8:

[1170] The terminal notifies the user of the quiz task and displays a quiz screen.

[1171] Step 9:

[1172] The user selects one of the options displayed on the quiz screen. For example, they select "Smartphone" and press the submit button.

[1173] Step 10:

[1174] The terminal transmits the selected answer to the server.

[1175] Step 11:

[1176] The server temporarily stores the received response data and starts data cleaning.

[1177] Step 12:

[1178] The server analyzes the received response data and removes inappropriate data and outliers.

[1179] Step 13:

[1180] The server adds the clean data to the AI's training dataset.

[1181] Step 14:

[1182] The server trains the AI ​​model based on the clean data.

[1183] Step 15:

[1184] The server generates a trained AI model as a generative AI plugin for a specific field.

[1185] Step 16:

[1186] The server provides this plugin to companies in specific fields or their own services via API.

[1187] Step 17:

[1188] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[1189] Step 18:

[1190] The terminal displays a notification to the user that points have been earned.

[1191] Example 1

[1192] 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."

[1193] Conventional ad-based services have the problem of being monotonous for users and low engagement. Furthermore, the revenue model based on ad viewing is unstable because it depends on the number of views. Furthermore, the accuracy of the collected data is low, making it difficult to obtain high-quality data, which is important for training generative AI models. A new system that can solve these issues and collect high-quality data while increasing user engagement is needed.

[1194] 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.

[1195] In this invention, the server includes means for inputting user authentication information and transmitting it from the terminal to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz-style task in exchange for watching an advertisement, means for the user to answer the quiz-style task and transmit the answer from the terminal to the server, means for the server to receive the answer data, analyze it to remove inappropriate data, and add the clean data to a training dataset, means for training a generative AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to organizations in a specific field via an API, and means for granting a reward each time a user completes a quiz-style task. This makes it possible to collect high-quality data and effectively train the generative AI model while increasing user engagement.

[1196] "User credentials" are the information a user enters to identify themselves and gain access, often consisting of a username and password.

[1197] A "terminal" is a device operated by a user to access the system, and includes, for example, a smartphone or a personal computer.

[1198] A "server" is a computer system that provides services over a network, and is a device that has the functions of data processing, storage, and responding to requests from clients.

[1199] A "login token" is authentication information that indicates that a user has been authenticated and is used to maintain a logged-in state for a certain period of time.

[1200] A "quiz-style task" is a quiz question that a user must answer, and is a task that includes multiple-choice questions and multiple options.

[1201] "Clean data" refers to high-quality data remaining after removing inappropriate data and outliers from the received response data.

[1202] A "training dataset" is a collection of data used to train a generative AI model, to which clean data has been added.

[1203] A "generative AI model" is a model of artificial intelligence that is trained based on collected data and algorithms used to perform specific tasks.

[1204] A "plug-in" is a software component with specific functionality that can be added to other software systems to extend their functionality.

[1205] "API" stands for Application Programming Interface, a standardized interface for communication between different software systems.

[1206] "Rewards" are points, special benefits, or other compensation that a user receives each time they complete a quiz-style task.

[1207] This invention is a system that collects high-quality data while increasing user engagement through a series of processes including user authentication, quiz-style task provision, data collection and analysis, generative AI model training, and a reward system. Below, we will explain the specific hardware and software usage methods and data processing and data calculation procedures.

[1208] User Authentication and Login

[1209] The device starts up, such as a smartphone or PC, and runs an application. The application displays a screen for entering user authentication information (username and password). The user enters this information and presses the login button. The device then sends the entered authentication information to the server using an HTTPS request.

[1210] The server checks the received authentication information against the database, and if authentication is successful, generates a JSON Web Token (JWT). This token is returned to the terminal, and a notification of successful login is displayed to the user, and the user is taken to the main screen.

[1211] Providing quiz-style tasks

[1212] After logging in, the server checks the user's usage history based on the user ID. If the usage history meets certain conditions, the server creates a quiz-style task instead of watching an advertisement. For example, the server generates a question such as "Which of the following products was the best-selling product in 2022?" along with options such as "smartphone," "tablet," "laptop," and "smartwatch."

[1213] The server sends this quiz task to the terminal, and the terminal notifies the user. The quiz screen is displayed on the terminal screen, and the user selects an answer from the displayed options.

[1214] User responses and data submission

[1215] The user selects an answer from the options and presses the submit button. The device sends the answer data to the server using an HTTPS request.

[1216] The server receives the response data and temporarily stores it. Next, it analyzes the data using Python's pandas library or similar to remove inappropriate data and outliers. The remaining clean data is added to the training dataset.

[1217] Training generative AI models

[1218] The server uses the clean data to train a generative AI model. For example, it trains a neural network model using frameworks such as TensorFlow or PyTorch to build a model that predicts product trends. This trained model is generated as a plugin.

[1219] The plugins are packaged in Docker containers and made available to field-specific organizations via an API, using a REST API that allows for easy integration with other systems.

[1220] User Rewards and Feedback

[1221] The server calculates the reward for the user who completes the quiz task. For example, 10 points are added to the user's account. The device notifies the user and displays a message saying "You've earned 10 points." The user can then check their point history in the app, which increases their motivation to participate in the next quiz task.

[1222] Specific examples

[1223] Example prompt: "Which of the following will be the best-selling products in 2022?" "Smartphones," "Tablets," "Laptops," "Smartwatches"

[1224] In this way, the system can collect high-quality data and effectively train generative AI models while increasing user engagement.

[1225] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1226] Processing Steps

[1227] Step 1: Launch the application and display the login screen

[1228] The device launches the application and displays a login screen, which displays fields for entering a username and password and a login button on the device screen.

[1229] Input: Login request from user

[1230] Output: A login screen appears

[1231] Step 2: Enter and submit user credentials

[1232] The user enters their username and password on the login screen. For example, the user enters "alice@example.com" and "password123" and clicks the Login button. The device sends this information to the server.

[1233] Input: The username and password entered by the user

[1234] Output: Authentication information sent to the server

[1235] Step 3: Validate credentials and generate tokens

[1236] The server checks the received authentication information against a database to verify the user's identity. If authentication is successful, the server generates a JSON Web Token (JWT). This token is used to maintain session information for the authenticated user.

[1237] Input: Authentication information received by the server

[1238] Output: Authentication result and generated JWT

[1239] Step 4: Login success notification and main screen display

[1240] The device saves the JWT received from the server and notifies the user that the login was successful. Specifically, it displays the message "Login successful" and transitions to the main screen.

[1241] Input: JWT sent from the server

[1242] Output: Login successful and main screen displayed

[1243] Step 5: Check usage history and generate quiz tasks

[1244] After logging in, the server checks the user's usage history. This history includes past quiz answer results and ad viewing history. If the server determines that it is time to view an ad based on the usage history, it generates a quiz-style task. For example, it could ask, "Which of the following products was the best-selling product in 2022?" with the options "smartphone," "tablet," "laptop," and "smartwatch."

[1245] Input: User usage history

[1246] Output: Generated quiz tasks

[1247] Step 6: Submitting and notifying quiz tasks

[1248] The server sends the generated quiz task to the terminal, which notifies the user of the quiz task and displays the quiz screen.

[1249] Input: Quiz task sent from the server

[1250] Output: The quiz screen shown to the user

[1251] Step 7: User responses and data submission

[1252] The user selects an answer from the options on the quiz screen and presses the send button. For example, they select "smartphone." The device then sends this answer data to the server.

[1253] Input: The user's selected answer

[1254] Output: Answer data sent to the server

[1255] Step 8: Receiving and analyzing response data

[1256] The server receives the submitted response data and temporarily stores it. It then analyzes the data using Python's pandas library and removes inappropriate data and outliers. Specifically, it performs a data cleaning process to generate clean data.

[1257] Input: Response data received by the server

[1258] Output: Cleaned data

[1259] Step 9: Add the clean data to the training dataset

[1260] The server adds the cleaned data to a training dataset, which is used to train a generative AI model.

[1261] Input: Cleaned data

[1262] Output: Updated training dataset

[1263] Step 10: Training the generative AI model

[1264] The server trains a generative AI model using the updated training dataset, for example, using TensorFlow or PyTorch to train a neural network model to create a model that performs a specific task (e.g., product trend prediction).

[1265] Input: Training dataset

[1266] Output: A trained generative AI model

[1267] Step 11: Generate and Provide the Plugin

[1268] The server generates trained generative AI models as plugins and packages them in Docker containers, and provides the plugins to organizations in specific fields via APIs, such as setting up endpoints using a REST API.

[1269] Input: A trained generative AI model

[1270] Output: Model provided as a plugin

[1271] Step 12: Calculating and notifying user rewards

[1272] The server calculates a reward for the user who has completed the quiz-style task, for example, adding 10 points to the user's account. The device notifies the user and displays a message saying "You have earned 10 points."

[1273] Input: Information about the completed quiz task

[1274] Output: Reward given to user and notification

[1275] (Application example 1)

[1276] 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."

[1277] Conventional data collection methods that rely on viewing advertisements have problems such as difficulty in data collection due to users' decreased willingness to view advertisements and the use of ad blockers. Another problem is that content recommendations based on individual users' interests are not possible, resulting in a poor user experience. To solve these problems, user-participation data collection methods and personalized content recommendation technologies are required.

[1278] 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.

[1279] In this invention, the server includes: means for inputting user authentication information and transmitting it to the server; means for the server to verify the authentication information and generate a login token if authenticated; means for providing authenticated users with a quiz task instead of viewing advertisements; means for the user to answer the quiz and transmit the answers to the server; means for the server to receive the answer data, remove inappropriate data, and add it to a training dataset; means for training an AI model based on the training dataset and generating the model as a plugin; means for providing the plugin to companies in a specific field via an API; means for rewarding users each time they complete a quiz task; means for generating and providing personalized content recommendation quizzes to authenticated users; and means for integrating and analyzing quiz answer data from multiple users to improve the performance of the AI ​​model. This enables user-participation data collection without advertisement viewing and personalized content recommendations based on individual user interests.

[1280] Creating definition statements

[1281] "User authentication information" refers to the identification information entered by a user to log into a system.

[1282] A "server" is a computing device that stores, processes, and manages data.

[1283] A "login token" is a temporary identifier that indicates an authenticated user's session and is used to authenticate subsequent communications.

[1284] A "quiz task" is a task in the form of a question provided to a user, and is a means for obtaining an answer from the user.

[1285] A "training dataset" is a collection of collected data used to train an AI model.

[1286] An "AI model" is an algorithm that analyzes data and performs a specific task.

[1287] A "plug-in" is a software module that is used to add specific functionality.

[1288] "API" stands for Application Program Interface, an interface for using functions between different software.

[1289] "Rewards" are incentives such as points or benefits that users can receive by completing quiz tasks.

[1290] A "personalized content recommendation quiz" is a quiz task that is individually provided based on the user's interests and concerns.

[1291] An "outlier" is a data point that is considered an anomaly in the collected data.

[1292] MODE FOR CARRYING OUT THE INVENTION

[1293] The present invention is a system that collects learning data for a generative AI by having users answer quiz-style tasks, and then trains an AI model based on this data. In this embodiment, the roles and specific operations of the user, terminal, and server are described.

[1294] User Authentication and Login

[1295] The user launches the smartphone app and logs in by entering their username and password on the login screen. This information is sent from the device to the server. The server references a database and verifies the authentication information. If authentication is successful, the server generates a login token and sends it to the device. The user is notified that the login was successful and is taken to the main screen.

[1296] Submitting a Quiz Task

[1297] After logging in, the server checks the user's usage history and generates a personalized quiz task. For example, it may ask a question such as "Which genre are you most interested in?" and provide the options "Movies," "Music," "Sports," and "News." The generated quiz task is sent to the terminal and notified to the user.

[1298] User responses and data submission

[1299] Users answer the quiz and press the submit button to send the answers from their device to the server. The server receives the answer data and removes inappropriate data and outliers. This clean data is added to the training dataset of the generative AI.

[1300] Data Processing and Validation

[1301] The server analyzes the collected response data and uses it as a learning dataset to train an AI model. This enables personalized content recommendations based on individual users' interests and preferences. The trained AI model is generated as a plugin and provided to companies in specific fields via API.

[1302] User Rewards and Feedback

[1303] Every time a user completes a quiz task, the server calculates the reward and adds it to the user's account. The terminal displays a notification to the user that points have been earned, encouraging them to participate in the next quiz task.

[1304] Hardware and software used

[1305] Hardware: Server, user's smartphone

[1306] Software: Flask (web application framework), SQLite (database), scikit-learn (machine learning library)

[1307] Specific examples

[1308] Consider the example of a prompt sentence in which a user answers a quiz on a smartphone app: "Which genre are you most interested in?"

[1309] Example prompt sentence:

[1310] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[1311] This allows users to enjoy content that interests them, and allows service providers to collect highly accurate data.

[1312] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1313] Program processing steps

[1314] Step 1:

[1315] The user launches the smartphone app and enters their username and password on the login screen. The entered authentication information is sent from the device to the server. The server receives this authentication information and authenticates the user by checking it against a database. If authentication is successful, the server generates a login token and sends it to the device. The device displays a notification to the user that login was successful and transitions to the main screen.

[1316] Input: Username, Password

[1317] Output: Login token, login success notification

[1318] Step 2:

[1319] The server checks the usage history of the logged-in user and generates a personalized quiz task based on that data. The content of the quiz task is, for example, a question such as "Which genre are you most interested in?" with multiple choices ("Movies," "Music," "Sports," "News"). The generated quiz task is sent to the terminal and notified to the user.

[1320] Input: User ID, usage history

[1321] Output: Quiz task

[1322] Step 3:

[1323] The user answers the quiz, selects one option, and presses the send button. This answer is sent from the device to the server.

[1324] Input: User's quiz answer

[1325] Output: Response data

[1326] Step 4:

[1327] The server receives the user's response data and performs analysis. Inappropriate data and outliers are removed, and the clean data is added to the training dataset of the generative AI.

[1328] Input: Response data

[1329] Output: Clean dataset

[1330] Step 5:

[1331] The server trains an AI model based on the clean data, which is then used to recommend personalized content. Once trained, the AI ​​model is generated as a plugin and made available to companies via API.

[1332] Input: Clean dataset

[1333] Output: Trained AI model, API plugin

[1334] Step 6:

[1335] When a user completes a quiz task, the server calculates the reward and adds it to the user's account. The device displays a notification to the user that points have been earned, encouraging them to participate in the quiz again.

[1336] Input: User ID, Quiz task completion information

[1337] Output: Reward points, notification

[1338] Example operation

[1339] Add specific examples of processing for each step.

[1340] Example of Step 1:

[1341] The user enters "alice@example.com" and "password123" into the smartphone app, the server authenticates, generates a login token, and returns it.

[1342] Example of Step 2:

[1343] The server checks the usage history of user "alice" and generates a quiz task asking "Which genre are you most interested in?" and provides the options "Movies," "Music," "Sports," and "News."

[1344] Example of Step 3:

[1345] The user selects "sports" and the answer data is sent to the server.

[1346] Example of Step 4:

[1347] The server analyzes the "Sports" responses, removes irrelevant data, and adds it to a clean dataset.

[1348] Example of Step 5:

[1349] The server uses the clean data collected to train an AI model and provides it to companies as a personalized content recommendation API.

[1350] Example of Step 6:

[1351] The server awards 10 points to user "alice" for completing the quiz task and displays a "Points Earned" notification.

[1352] Prompt Sentence Examples

[1353] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[1354] 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.

[1355] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state, by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. Specific embodiments of this system are described below.

[1356] System Overview

[1357] The system consists of the following main components:

[1358] 1. User authentication system

[1359] 2. Quiz Task Providing System

[1360] 3. Response data processing system

[1361] 4. Emotion Engine

[1362] 5. Learning Dataset Management System

[1363] 6. AI Model Training System

[1364] 7. Plugin Creation and Distribution System

[1365] 8. Reward Management System

[1366] User Authentication and Login

[1367] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1368] The terminal transmits the entered user authentication information to the server.

[1369] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1370] The device will notify the user that login was successful and transition to the main screen.

[1371] Submitting a Quiz Task

[1372] After logging in, the server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[1373] The server generates a quiz task such as "Which of the following products will be the best-selling in 2022?" and presents multiple options (e.g., smartphone, tablet, laptop, smartwatch).

[1374] The server sends the generated quiz task to the terminal.

[1375] The terminal notifies the user of the quiz task and displays a quiz screen.

[1376] User responses and data submission

[1377] The user selects one of the options displayed on the quiz screen and presses the submit button (e.g., selects smartphone).

[1378] The terminal transmits the selected answer to the server.

[1379] Data Processing and Sentiment Analysis

[1380] The server temporarily stores the received response data and starts data cleaning.

[1381] The server analyzes the received response data and removes inappropriate data and outliers.

[1382] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[1383] Learning dataset and AI model training

[1384] The server adds the clean data and emotion data to the AI's training dataset.

[1385] The server trains the AI ​​model based on the clean data and emotion data.

[1386] For example, quiz answers and emotional data collected from multiple users can be fed into an AI model to improve prediction accuracy.

[1387] Plugin generation and provision

[1388] The server generates a trained AI model as a generative AI plugin for a specific field.

[1389] The server provides the generated plugins to companies in specific fields or their own services via API.

[1390] Compensation management

[1391] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[1392] The terminal displays a notification to the user that points have been earned.

[1393] Specific examples

[1394] 1. User authentication and login details example

[1395] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[1396] The terminal sends this information to the server, which checks the information against a database.

[1397] If the authentication is successful, the server generates a token and returns it to the terminal.

[1398] The device will notify you that the login was successful and will transition to the main screen.

[1399] 2. Example of Quiz Task Submission Details

[1400] The server checks the usage history and emotional state of user "alice" and determines that it is time for her to watch an advertisement.

[1401] The emotion engine analyzes "Alice's" emotions in real time and determines, for example, that she is "relaxed."

[1402] The server generates a quiz task asking, "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[1403] The terminal displays the quiz and notifies the user.

[1404] 3. Detailed example of user responses and data submission

[1405] The user selects "smartphone" and presses the send button.

[1406] The terminal transmits this response data to the server.

[1407] The server analyzes the received response data and removes inappropriate data.

[1408] 4. Detailed examples of data processing and sentiment analysis

[1409] The server adds "smartphone" to the training dataset as clean data.

[1410] The emotion engine collects emotional data (e.g., "joy") from users' responses and adds it to the learning dataset.

[1411] 5. Detailed example of creating and providing a plugin

[1412] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[1413] The server provides this plugin to companies and their own services via API.

[1414] 6. Detailed Example of User Rewards and Feedback

[1415] The server calculates a reward of 10 points for "alice" and adds it to her account.

[1416] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[1417] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions to generate more effective learning data and improve the performance of the generation AI.

[1418] The processing flow will be explained below.

[1419] Step 1:

[1420] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1421] Step 2:

[1422] The terminal transmits the entered user authentication information to the server.

[1423] Step 3:

[1424] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1425] Step 4:

[1426] The device will notify the user that login was successful and transition to the main screen.

[1427] Step 5:

[1428] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[1429] Step 6:

[1430] The emotion engine analyzes the user's emotions and adjusts the content and difficulty of the quiz task according to their state. For example, if the user is relaxed, a more difficult quiz is presented, and if the user is stressed, an easier quiz is presented.

[1431] Step 7:

[1432] The server generates a quiz task that asks the question, "Which of the following products will be the best-selling in 2022?" and provides several options (e.g., smartphone, tablet, laptop, smartwatch).

[1433] Step 8:

[1434] The server sends the generated quiz task to the terminal.

[1435] Step 9:

[1436] The terminal notifies the user of the quiz task and displays a quiz screen.

[1437] Step 10:

[1438] The user selects one of the options displayed on the quiz screen, for example, "smartphone," and presses the send button.

[1439] Step 11:

[1440] The terminal transmits the selected answer to the server.

[1441] Step 12:

[1442] The server temporarily stores the received response data and starts cleaning the data.

[1443] Step 13:

[1444] The server analyzes the received response data and removes inappropriate data and outliers.

[1445] Step 14:

[1446] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[1447] Step 15:

[1448] The server adds the clean data and emotion data to the AI's training dataset.

[1449] Step 16:

[1450] The server trains the AI ​​model based on the clean data and emotion data.

[1451] Step 17:

[1452] The server generates a trained AI model as a generative AI plugin for a specific field.

[1453] Step 18:

[1454] The server provides the generated plugins to companies in specific fields or their own services via API.

[1455] Step 19:

[1456] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[1457] Step 20:

[1458] The terminal displays a notification to the user that points have been earned.

[1459] Example 2

[1460] 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."

[1461] Conventional ad viewing formats require users to passively watch ads, resulting in poor user experience and limited advertising effectiveness. Furthermore, because interactive tasks that take into account the user's emotional state are not provided, it is difficult to effectively collect data based on the user's interests. Furthermore, there are issues with quality control of collected data and optimizing training datasets, which can lead to the inclusion of inappropriate data.

[1462] 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 inputting user authentication information and transmitting it to a device on the network; means for the device on the network to verify the authentication information and generate a login token if authenticated; and means for providing the authenticated user with a quiz task instead of viewing an advertisement. This enables the user's advertisement viewing experience to be replaced with an active quiz task. The server also includes means for the user to answer the quiz and transmit the answer to a device on the network; and means for the device on the network to receive the answer data, clean the data, and add it to a training dataset. This improves quality control of collected data and enables the generation of an appropriate training dataset. The server also includes means for training an artificial intelligence model based on the training dataset and generating the model as a plugin, means for providing the plugin to entities in a specific field via an API, and means for granting a reward each time a user completes a quiz task. This improves the performance of the AI ​​model and increases user motivation to participate.

[1463] A "networked device" is an electronic device such as a computer, smartphone, or tablet that can communicate over the Internet or other networks.

[1464] "Authentication Information" refers to the username, password, and other authentication means entered by a user when logging into a system.

[1465] A "login token" is a unique identifier or security token that is generated when a server successfully authenticates a user and is used to maintain the user's login session.

[1466] A "quiz task" is a question-type task in which the user can participate interactively, and in which the user selects the correct answer from multiple options.

[1467] "Data cleaning" is the process of automatically removing outliers and irrelevant data from received data, a process that aims to improve data quality.

[1468] A "training dataset" is a collection of data used to train an artificial intelligence model, including clean data and sentiment data.

[1469] An "artificial intelligence model" is a machine learning algorithm that is trained on a collected dataset and is a computer program to automatically perform a specific task.

[1470] A "plugin" is a software module that adds specific functionality or services to other systems or applications via an API.

[1471] "Entity" is a general term that refers to a legal entity, company, or organization that provides business or services in a particular field.

[1472] "API" is an abbreviation for Application Programming Interface, an interface for sharing functions between different software applications.

[1473] "Rewards" refer to incentives such as points or benefits that are given to users each time they complete a quiz task.

[1474] System Overview

[1475] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. The system is implemented using devices (servers and terminals) on a network.

[1476] Hardware and software used

[1477] Hardware: Servers, devices (computers, smartphones, tablets)

[1478] Software: User authentication system, quiz task provision system, answer data processing system, emotion engine, learning dataset management system, AI model training system, reward management system

[1479] User Authentication and Login

[1480] The user starts the application using the device and the login screen is displayed. The user enters a username and password and sends them to the server. The server compares the received authentication information with the database, and if authentication is successful, generates a login token and sends it back to the device. The device displays a notification to the user that the login was successful and transitions to the main screen.

[1481] Submitting a Quiz Task

[1482] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz task based on that. The server generates quiz tasks such as "Which of the following products was the best-selling in 2022?" and presents multiple options. The server sends the generated quiz task to the device, and the device notifies the user of the quiz task and displays the quiz screen.

[1483] User responses and data submission

[1484] The user selects one answer from the options displayed on the quiz screen and presses the send button. The device then sends the selected answer to the server.

[1485] Data Processing and Sentiment Analysis

[1486] The server temporarily stores the received answer data and starts data cleaning. The server analyzes the received answer data and removes inappropriate data and outliers. The emotion engine collects emotional data (e.g., joy, surprise, irritation) from the user when answering the quiz and adds it to the training dataset.

[1487] Learning dataset and AI model training

[1488] The server adds the cleaned response data and sentiment data to the AI ​​training dataset, which the server uses to train the AI ​​model and improve its prediction accuracy.

[1489] Plugin generation and provision

[1490] The server generates the trained AI model as a generative AI plugin for a specific domain, and provides the generated plugin to entities in the specific domain via an API.

[1491] Compensation management

[1492] The server calculates the reward (points or rewards) for the user who completed the quiz task and adds it to the user's account. The terminal displays a notification to the user that points have been earned.

[1493] Specific examples

[1494] 1. Example of user authentication:

[1495] The user launches the app, enters "alice@example.com" and "password123," and presses the login button. The device sends the information to the server, which checks the information in the database. If authentication is successful, the server generates a token and sends it back to the device. The device notifies the user that "Login was successful" and returns to the main screen.

[1496] 2. Example of a quiz task:

[1497] The server checks the usage history and emotional state of user "alice," which is "relaxed." The server generates a quiz task asking, "Which of the following products was the best-selling product in 2022?" and presents the options "smartphone," "tablet," "laptop," and "smartwatch." The device displays the quiz screen and notifies the user.

[1498] 3. Example prompt:

[1499] "Which of the following products will be the best-selling in 2022?"

[1500] "Predict what kind of ad would be best based on the emotional data of the user when they answer the quiz."

[1501] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions, thereby generating more effective training data and improving the performance of the generation AI.

[1502] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1503] Step 1:

[1504] A user starts an application using a terminal and a login screen is displayed. The user enters a username and password. The terminal sends the entered authentication information (username and password) to the server. The username and password are used as input data, and that data is sent to the server.

[1505] Step 2:

[1506] The server compares the received authentication information with the database and generates a login token if authentication is successful. It searches the database for matching authentication information and obtains the result of successful authentication as output. If authentication is successful, a login token is generated and the server sends that login token to the terminal. The output is the login token if successful. As a specific example, the server compares the username "alice@example.com" with the password "password123" and generates a login token if successful.

[1507] Step 3:

[1508] Based on the login token received by the terminal from the server, the terminal displays a notification of successful login to the user and transitions to the main screen. The login token received from the server is used as input. As output, the terminal displays a "login successful" notification to the user and transitions to the main screen.

[1509] Step 4:

[1510] The server checks the usage history and emotional state based on the user ID. As input data, it compares past usage history and real-time emotional state data based on the user ID. As output, it obtains the confirmed usage history and the emotion analysis results. For example, the emotion engine analyzes the emotion of user "alice" as "relaxed."

[1511] Step 5:

[1512] The server generates quiz tasks by adjusting the difficulty and type of the quiz task based on the emotional state. The appropriate quiz task is determined based on the input data of the emotion analysis results and the user's usage history. The generated quiz task is obtained as output. As a specific example, the server generates a quiz question such as, "Which of the following products was the best-selling product in 2022?"

[1513] Step 6:

[1514] The server sends the generated quiz task to the terminal, and the terminal notifies the user of the quiz task and displays the quiz screen. As input, the terminal receives the quiz task sent from the server. As output, the quiz screen is displayed to the user. The terminal displays the quiz screen and notifies the user to "answer the quiz."

[1515] Step 7:

[1516] The user selects one of the options displayed on the quiz screen and presses the send button. The answer data selected by the user is collected as input data. The selected answer data is obtained as output. For example, the user selects "smartphone" and presses the send button.

[1517] Step 8:

[1518] The device sends the selected answer data to the server. As input, it receives the answer data selected by the user. As output, it obtains the answer data to be sent to the server. The device sends the answer data "smartphone" to the server.

[1519] Step 9:

[1520] The server temporarily stores the received response data and begins data cleaning. Inappropriate data and outliers are removed. The received response data is used as input. Cleaned, clean data is obtained as output. The server stores the "smartphone" as clean data and removes inappropriate data.

[1521] Step 10:

[1522] The emotion engine collects emotional data when users answer quizzes and adds it to the training dataset. It uses real-time emotional data when users answer quizzes as input. It obtains the collected emotional data as output. For example, it collects the user's emotion when answering a quiz as "joy."

[1523] Step 11:

[1524] The server adds the cleaned response data and emotion data to the AI's training dataset. The cleaned data and emotion data are used as input. The output is an updated training dataset.

[1525] Step 12:

[1526] The server trains the AI ​​model based on the training dataset to improve prediction accuracy. The training dataset is used as input. The output is a trained AI model.

[1527] Step 13:

[1528] The server generates a trained AI model as a generative AI plugin for a specific field. The trained AI model is used as input. The generated plugin is obtained as output.

[1529] Step 14:

[1530] The server provides the generated plugin to a domain-specific entity via an API, uses the generated plugin as input, and sends the provided plugin to the domain-specific entity as output.

[1531] Step 15:

[1532] The server calculates a reward for a user who completes a quiz task and adds it to the user's account. As input, it uses the quiz task completion data. As output, the calculated reward is added to the user's account. The server calculates 10 points for "alice" and adds it to the account.

[1533] Step 16:

[1534] The terminal displays a notification to the user that a reward has been earned. As input, it receives the reward data sent from the server. As output, it displays a notification to the user that a reward has been earned. The terminal notifies the user that "points have been earned," motivating the user to participate in the next quiz task.

[1535] (Application example 2)

[1536] 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."

[1537] Conventional advertising viewing methods have made it difficult to effectively deliver ads to users and accurately measure ad performance. Furthermore, they have been unable to optimize ads based on the user's emotional state, leaving a lack of means to improve the user's advertising experience. The present invention aims to achieve more effective and personalized ad delivery by measuring advertising effectiveness through quiz-style tasks and analyzing the user's emotional state.

[1538] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1539] In this invention, the server includes means for inputting user authentication information and transmitting it to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz task instead of watching an advertisement, means for the user to answer the quiz and transmit the answer to the server, means for the server to receive the answer data, remove inappropriate data, and add the answer to a training dataset, means for the server to analyze the user's emotional data when answering the quiz and add the data to the training dataset, means for training an AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to companies in specific fields via an API, and means for awarding a reward each time a user completes a quiz task. This makes it possible to analyze the user's emotional state and provide more effective advertisement delivery and a personalized user experience.

[1540] "User authentication information" is information for identifying users and controlling access.

[1541] A "login token" is a temporary session identifier for an authenticated user to access a system.

[1542] A "quiz task" is a challenge presented to a user consisting of a series of questions and multiple-choice options.

[1543] "Answer data" is information about the answer selected by the user for the quiz task.

[1544] "Inappropriate data" is data that is not worthy or reliable for the system to include in the training dataset.

[1545] "Emotion data" is data that indicates the user's psychological state and emotions.

[1546] A "training dataset" is a set of clean data and sentiment data used to train an AI model.

[1547] An "AI model" is a software architecture that is trained to solve a specific problem based on machine learning algorithms.

[1548] A "plug-in" is a software component that extends or adds specific functionality.

[1549] "API" stands for Application Programming Interface, an interface for exchanging data and functions between different software systems.

[1550] "Rewards" are incentives such as points or benefits that users receive by completing quiz tasks.

[1551] The present invention is a system that combines a user-generated AI with learning data provided by the AI ​​through quiz-style tasks instead of watching advertisements, and an emotion engine that recognizes the user's emotional state. To realize this system, the following steps must be performed:

[1552] First, the device collects user authentication information and sends it to the server. The server verifies the authentication information, and if authentication is successful, generates a login token and returns it to the device. Once authentication is complete, the server provides the user with a quiz-style task instead of watching an advertisement. This quiz task is customized based on the user's usage history and real-time emotional state.

[1553] When a user answers a quiz, the device sends the answer data to the server. The server analyzes the received answer data, removes inappropriate data and outliers, and adds the clean data to the training dataset. At the same time, the server uses an emotion engine to analyze the user's emotional data when answering the quiz and adds that data to the training dataset.

[1554] The server then trains a generative AI model based on this learning dataset. The resulting AI model is then generated as a generative AI plugin for a specific field. This plugin is then provided to companies in that field or to their own services via API.

[1555] Through this process, the server will reward users each time they complete a quiz task. The collected clean data and emotional data will improve the accuracy of the AI ​​model, allowing companies to provide more effective advertising and services.

[1556] The hardware and software used are as follows:

[1557] Smartphone: The device through which the user accesses and answers the quiz tasks.

[1558] Python: A programming language used to develop each component of the system.

[1559] Emotion analysis model: An AI model for analyzing a user's emotional state.

[1560] Server: Validates authentication information, provides quiz tasks, processes answer and sentiment data, trains AI models, generates and provides plugins, and grants rewards.

[1561] Examples:

[1562] For example, imagine a user using an app. The user logs in to the app using a smartphone. After successfully logging in, the server generates a quiz question: "What was the best-selling product in 2022?" and displays it on the smartphone. The user selects "smartphone" and submits the answer. The server analyzes the answer data and uses an emotion engine to collect the user's emotional data at the time of answering as "joy." Finally, the user receives a reward of 10 points.

[1563] Example prompts to input to a generative AI model:

[1564] "Answer the following question: Which of the following products was your top-selling item in 2022?"

[1565] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1566] Step 1:

[1567] The terminal obtains the user authentication information and sends it to the server.

[1568] Enter your username and password.

[1569] Output: Sends authentication information to the server.

[1570] How it works: A user launches an application on their smartphone and enters their username and password into the login screen. The device then securely transmits this information to the server using the HTTPS protocol.

[1571] Step 2:

[1572] The server verifies the authentication information and generates a login token if authenticated.

[1573] Input: The username and password submitted.

[1574] Output: Authentication result and login token.

[1575] What happens: The server checks the user information stored in the database against the authentication information sent. If authentication is successful, the server generates a temporary login token and sends it back to the device.

[1576] Step 3:

[1577] The server provides the authenticated user with a quiz task in exchange for viewing an advertisement.

[1578] Input: User ID and login token, usage history, emotional state.

[1579] Output: A customized quiz task.

[1580] How it works: The server checks the usage history and real-time emotional state of the authenticated user and generates a quiz task based on that. For example, it provides a question like "What will be the best-selling product in 2022?" with options (smartphone, tablet, laptop, smartwatch).

[1581] Step 4:

[1582] The terminal presents the generated quiz task to the user, and the user answers the quiz.

[1583] Input: A customized quiz task.

[1584] Output: The user's answer.

[1585] Specific operation: The device notifies the user of the quiz task, displays the question and options, and the user selects one and presses the answer button.

[1586] Step 5:

[1587] The terminal transmits the user's response data to the server.

[1588] Input: The user's answer.

[1589] Output: Send the response data to the server.

[1590] Specific operation: When the user selects an answer to the quiz and presses the send button, the device sends the answer data to the server.

[1591] Step 6:

[1592] The server analyzes the received response data and removes inappropriate data.

[1593] Input: User response data.

[1594] Output: Clean data.

[1595] Specific operation: The server validates the received response data, detects and removes inappropriate data and outliers, and only reliable data is added to the training dataset.

[1596] Step 7:

[1597] The server uses an emotion engine to analyze the emotion data of the user when answering the quiz.

[1598] Input: User response data.

[1599] Output: User emotion data.

[1600] Specific operation: The server uses an emotion engine to analyze the user's facial expressions and voice patterns when answering the quiz, and identifies emotions such as joy, surprise, and irritation.

[1601] Step 8:

[1602] The server adds the clean data and the emotion data to the training dataset.

[1603] Input: Clean data, sentiment data.

[1604] Output: Added to the training dataset.

[1605] Specific operation: The server stores the clean data and analyzed emotion data in a learning dataset and uses it as training data for the generative AI model.

[1606] Step 9:

[1607] The server trains a generative AI model based on the learning dataset and generates the model as a plugin.

[1608] Input: The training dataset.

[1609] Output: The trained AI model plugin.

[1610] What it does: The server runs machine learning algorithms on a training dataset to train a generative AI model for use in a specific domain, and outputs the completed model in the form of a plugin.

[1611] Step 10:

[1612] The server provides the generated AI plugins to companies in specific fields via API.

[1613] Input: A trained AI model plugin.

[1614] Output: Provided via API.

[1615] How it works: The server provides trained AI model plugins to companies in specific fields via an API interface, which then use the plugin's functions in their own systems.

[1616] Step 11:

[1617] The server awards a reward each time a user completes a quiz task.

[1618] Input: User ID.

[1619] Output: Reward points.

[1620] How it works: The server monitors the user's quiz task completion status and adds a set reward point to the user's account every time the task is completed. This reward motivates the user.

[1621] 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.

[1622] 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.

[1623] 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.

[1624] [Fourth embodiment]

[1625] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1626] 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.

[1627] 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).

[1628] 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.

[1629] 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.

[1630] 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).

[1631] 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.

[1632] 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.

[1633] 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.

[1634] 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.

[1635] 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.

[1636] 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.

[1637] 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."

[1638] The present invention is a system that provides quiz-style tasks in exchange for watching advertisements, and collects learning data for a generation AI by having users answer the quiz tasks, and trains an AI model based on this data. The following describes in natural language the roles and specific operations of the user, terminal, and server in an embodiment of the present invention.

[1639] 1. User authentication and login

[1640] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1641] The terminal transmits the entered user authentication information to the server.

[1642] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1643] The device will notify the user that login was successful and transition to the main screen.

[1644] 2. Submit the quiz task

[1645] After logging in, the server checks the usage history based on the user ID.

[1646] If the server determines that it is time to watch an ad, it generates a quiz-style learning task, such as "Which of the following products will be the best-selling in 2022?" with multiple choices.

[1647] The server sends the generated quiz task to the terminal.

[1648] The terminal notifies the user of the quiz task and displays a quiz screen.

[1649] 3. User responses and data submission

[1650] The user selects one of the options displayed (e.g., "smartphone," "tablet," "laptop," or "smartwatch") and presses the send button.

[1651] The terminal transmits the selected answer to the server.

[1652] The server temporarily stores the received response data and proceeds to the next process.

[1653] 4. Data processing and validation

[1654] The server analyzes the received response data and removes inappropriate data and outliers.

[1655] The remaining clean data is added to the AI ​​training dataset.

[1656] For example, answer data collected from multiple users to the question, "What was the best-selling product in 2022?" is integrated and provided to the generative AI as training data.

[1657] 5. Creating and Providing Plugins

[1658] The server trains the AI ​​model using the clean data.

[1659] A trained AI model can be plugged into a specific field (e.g., "product trend prediction").

[1660] The server provides the generated plugins to companies in specific fields or their own services via API.

[1661] 6. User Rewards and Feedback

[1662] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to their accounts.

[1663] The terminal displays a notification to the user that points have been earned.

[1664] Users can check the points and rewards they have earned, which will motivate them to take part in the next quiz task.

[1665] Specific examples

[1666] 1. User authentication and login details example

[1667] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[1668] The terminal sends this information to the server, which checks the information against a database.

[1669] If the authentication is successful, the server generates a token and returns it to the terminal.

[1670] The device will notify you that the login was successful and will transition to the main screen.

[1671] 2. Example of Quiz Task Submission Details

[1672] The server checks the usage history of user "alice" and determines that it is time to view an advertisement.

[1673] The server generates a quiz task: "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[1674] The terminal displays the quiz and notifies the user.

[1675] 3. Detailed example of user responses and data submission

[1676] The user selects "smartphone" and presses the send button.

[1677] The terminal transmits this response data to the server.

[1678] The server analyzes the received response data and removes inappropriate data.

[1679] 4. Detailed examples of data processing and validation

[1680] The server adds "smartphone" to the training dataset as clean data.

[1681] 5. Detailed example of creating and providing a plugin

[1682] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[1683] The server provides this plugin to companies and their own services via API.

[1684] 6. Detailed Example of User Rewards and Feedback

[1685] The server calculates a reward of 10 points for "alice" and adds it to her account.

[1686] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[1687] This allows users to participate in quizzes instead of watching ads, and also improves the performance of the generative AI.

[1688] The processing flow will be explained below.

[1689] Step 1:

[1690] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1691] Step 2:

[1692] The terminal transmits the entered user authentication information to the server.

[1693] Step 3:

[1694] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1695] Step 4:

[1696] The device will notify the user that login was successful and transition to the main screen.

[1697] Step 5:

[1698] The server checks the usage history based on the user ID and determines whether it is time to watch an advertisement.

[1699] Step 6:

[1700] If the server determines that an ad is required, it generates a quiz task. For example, it prepares a question such as, "Which of the following products was the best-selling in 2022?" and presents several options.

[1701] Step 7:

[1702] The server sends the generated quiz task to the terminal.

[1703] Step 8:

[1704] The terminal notifies the user of the quiz task and displays a quiz screen.

[1705] Step 9:

[1706] The user selects one of the options displayed on the quiz screen. For example, they select "Smartphone" and press the submit button.

[1707] Step 10:

[1708] The terminal transmits the selected answer to the server.

[1709] Step 11:

[1710] The server temporarily stores the received response data and starts data cleaning.

[1711] Step 12:

[1712] The server analyzes the received response data and removes inappropriate data and outliers.

[1713] Step 13:

[1714] The server adds the clean data to the AI's training dataset.

[1715] Step 14:

[1716] The server trains the AI ​​model based on the clean data.

[1717] Step 15:

[1718] The server generates a trained AI model as a generative AI plugin for a specific field.

[1719] Step 16:

[1720] The server provides this plugin to companies in specific fields or their own services via API.

[1721] Step 17:

[1722] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[1723] Step 18:

[1724] The terminal displays a notification to the user that points have been earned.

[1725] Example 1

[1726] 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."

[1727] Conventional ad-based services have the problem of being monotonous for users and low engagement. Furthermore, the revenue model based on ad viewing is unstable because it depends on the number of views. Furthermore, the accuracy of the collected data is low, making it difficult to obtain high-quality data, which is important for training generative AI models. A new system that can solve these issues and collect high-quality data while increasing user engagement is needed.

[1728] 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.

[1729] In this invention, the server includes means for inputting user authentication information and transmitting it from the terminal to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz-style task in exchange for watching an advertisement, means for the user to answer the quiz-style task and transmit the answer from the terminal to the server, means for the server to receive the answer data, analyze it to remove inappropriate data, and add the clean data to a training dataset, means for training a generative AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to organizations in a specific field via an API, and means for granting a reward each time a user completes a quiz-style task. This makes it possible to collect high-quality data and effectively train the generative AI model while increasing user engagement.

[1730] "User credentials" are the information a user enters to identify themselves and gain access, often consisting of a username and password.

[1731] A "terminal" is a device operated by a user to access the system, and includes, for example, a smartphone or a personal computer.

[1732] A "server" is a computer system that provides services over a network, and is a device that has the functions of data processing, storage, and responding to requests from clients.

[1733] A "login token" is authentication information that indicates that a user has been authenticated and is used to maintain a logged-in state for a certain period of time.

[1734] A "quiz-style task" is a quiz question that a user must answer, and is a task that includes multiple-choice questions and multiple options.

[1735] "Clean data" refers to high-quality data remaining after removing inappropriate data and outliers from the received response data.

[1736] A "training dataset" is a collection of data used to train a generative AI model, to which clean data has been added.

[1737] A "generative AI model" is a model of artificial intelligence that is trained based on collected data and algorithms used to perform specific tasks.

[1738] A "plug-in" is a software component with specific functionality that can be added to other software systems to extend their functionality.

[1739] "API" stands for Application Programming Interface, a standardized interface for communication between different software systems.

[1740] "Rewards" are points, special benefits, or other compensation that a user receives each time they complete a quiz-style task.

[1741] This invention is a system that collects high-quality data while increasing user engagement through a series of processes including user authentication, quiz-style task provision, data collection and analysis, generative AI model training, and a reward system. Below, we will explain the specific hardware and software usage methods and data processing and data calculation procedures.

[1742] User Authentication and Login

[1743] The device starts up, such as a smartphone or PC, and runs an application. The application displays a screen for entering user authentication information (username and password). The user enters this information and presses the login button. The device then sends the entered authentication information to the server using an HTTPS request.

[1744] The server checks the received authentication information against the database, and if authentication is successful, generates a JSON Web Token (JWT). This token is returned to the terminal, and a notification of successful login is displayed to the user, and the user is taken to the main screen.

[1745] Providing quiz-style tasks

[1746] After logging in, the server checks the user's usage history based on the user ID. If the usage history meets certain conditions, the server creates a quiz-style task instead of watching an advertisement. For example, the server generates a question such as "Which of the following products was the best-selling product in 2022?" along with options such as "smartphone," "tablet," "laptop," and "smartwatch."

[1747] The server sends this quiz task to the terminal, and the terminal notifies the user. The quiz screen is displayed on the terminal screen, and the user selects an answer from the displayed options.

[1748] User responses and data submission

[1749] The user selects an answer from the options and presses the submit button. The device sends the answer data to the server using an HTTPS request.

[1750] The server receives the response data and temporarily stores it. Next, it analyzes the data using Python's pandas library or similar to remove inappropriate data and outliers. The remaining clean data is added to the training dataset.

[1751] Training generative AI models

[1752] The server uses the clean data to train a generative AI model. For example, it trains a neural network model using frameworks such as TensorFlow or PyTorch to build a model that predicts product trends. This trained model is generated as a plugin.

[1753] The plugins are packaged in Docker containers and made available to field-specific organizations via an API, using a REST API that allows for easy integration with other systems.

[1754] User Rewards and Feedback

[1755] The server calculates the reward for the user who completes the quiz task. For example, 10 points are added to the user's account. The device notifies the user and displays a message saying "You've earned 10 points." The user can then check their point history in the app, which increases their motivation to participate in the next quiz task.

[1756] Specific examples

[1757] Example prompt: "Which of the following will be the best-selling products in 2022?" "Smartphones," "Tablets," "Laptops," "Smartwatches"

[1758] In this way, the system can collect high-quality data and effectively train generative AI models while increasing user engagement.

[1759] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1760] Processing Steps

[1761] Step 1: Launch the application and display the login screen

[1762] The device launches the application and displays a login screen, which displays fields for entering a username and password and a login button on the device screen.

[1763] Input: Login request from user

[1764] Output: A login screen appears

[1765] Step 2: Enter and submit user credentials

[1766] The user enters their username and password on the login screen. For example, the user enters "alice@example.com" and "password123" and clicks the Login button. The device sends this information to the server.

[1767] Input: The username and password entered by the user

[1768] Output: Authentication information sent to the server

[1769] Step 3: Validate credentials and generate tokens

[1770] The server checks the received authentication information against a database to verify the user's identity. If authentication is successful, the server generates a JSON Web Token (JWT). This token is used to maintain session information for the authenticated user.

[1771] Input: Authentication information received by the server

[1772] Output: Authentication result and generated JWT

[1773] Step 4: Login success notification and main screen display

[1774] The device saves the JWT received from the server and notifies the user that the login was successful. Specifically, it displays the message "Login successful" and transitions to the main screen.

[1775] Input: JWT sent from the server

[1776] Output: Login successful and main screen displayed

[1777] Step 5: Check usage history and generate quiz tasks

[1778] After logging in, the server checks the user's usage history. This history includes past quiz answer results and ad viewing history. If the server determines that it is time to view an ad based on the usage history, it generates a quiz-style task. For example, it could ask, "Which of the following products was the best-selling product in 2022?" with the options "smartphone," "tablet," "laptop," and "smartwatch."

[1779] Input: User usage history

[1780] Output: Generated quiz tasks

[1781] Step 6: Submitting and notifying quiz tasks

[1782] The server sends the generated quiz task to the terminal, which notifies the user of the quiz task and displays the quiz screen.

[1783] Input: Quiz task sent from the server

[1784] Output: The quiz screen shown to the user

[1785] Step 7: User responses and data submission

[1786] The user selects an answer from the options on the quiz screen and presses the send button. For example, they select "smartphone." The device then sends this answer data to the server.

[1787] Input: The user's selected answer

[1788] Output: Answer data sent to the server

[1789] Step 8: Receiving and analyzing response data

[1790] The server receives the submitted response data and temporarily stores it. It then analyzes the data using Python's pandas library and removes inappropriate data and outliers. Specifically, it performs a data cleaning process to generate clean data.

[1791] Input: Response data received by the server

[1792] Output: Cleaned data

[1793] Step 9: Add the clean data to the training dataset

[1794] The server adds the cleaned data to a training dataset, which is used to train a generative AI model.

[1795] Input: Cleaned data

[1796] Output: Updated training dataset

[1797] Step 10: Training the generative AI model

[1798] The server trains a generative AI model using the updated training dataset, for example, using TensorFlow or PyTorch to train a neural network model to create a model that performs a specific task (e.g., product trend prediction).

[1799] Input: Training dataset

[1800] Output: A trained generative AI model

[1801] Step 11: Generate and Provide the Plugin

[1802] The server generates trained generative AI models as plugins and packages them in Docker containers, and provides the plugins to organizations in specific fields via APIs, such as setting up endpoints using a REST API.

[1803] Input: A trained generative AI model

[1804] Output: Model provided as a plugin

[1805] Step 12: Calculating and notifying user rewards

[1806] The server calculates a reward for the user who has completed the quiz-style task, for example, adding 10 points to the user's account. The device notifies the user and displays a message saying "You have earned 10 points."

[1807] Input: Information about the completed quiz task

[1808] Output: Reward given to user and notification

[1809] (Application example 1)

[1810] 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."

[1811] Conventional data collection methods that rely on viewing advertisements have problems such as difficulty in data collection due to users' decreased willingness to view advertisements and the use of ad blockers. Another problem is that content recommendations based on individual users' interests are not possible, resulting in a poor user experience. To solve these problems, user-participation data collection methods and personalized content recommendation technologies are required.

[1812] 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.

[1813] In this invention, the server includes: means for inputting user authentication information and transmitting it to the server; means for the server to verify the authentication information and generate a login token if authenticated; means for providing authenticated users with a quiz task instead of viewing advertisements; means for the user to answer the quiz and transmit the answers to the server; means for the server to receive the answer data, remove inappropriate data, and add it to a training dataset; means for training an AI model based on the training dataset and generating the model as a plugin; means for providing the plugin to companies in a specific field via an API; means for rewarding users each time they complete a quiz task; means for generating and providing personalized content recommendation quizzes to authenticated users; and means for integrating and analyzing quiz answer data from multiple users to improve the performance of the AI ​​model. This enables user-participation data collection without advertisement viewing and personalized content recommendations based on individual user interests.

[1814] Creating definition statements

[1815] "User authentication information" refers to the identification information entered by a user to log into a system.

[1816] A "server" is a computing device that stores, processes, and manages data.

[1817] A "login token" is a temporary identifier that indicates an authenticated user's session and is used to authenticate subsequent communications.

[1818] A "quiz task" is a task in the form of a question provided to a user, and is a means for obtaining an answer from the user.

[1819] A "training dataset" is a collection of collected data used to train an AI model.

[1820] An "AI model" is an algorithm that analyzes data and performs a specific task.

[1821] A "plug-in" is a software module that is used to add specific functionality.

[1822] "API" stands for Application Program Interface, an interface for using functions between different software.

[1823] "Rewards" are incentives such as points or benefits that users can receive by completing quiz tasks.

[1824] A "personalized content recommendation quiz" is a quiz task that is individually provided based on the user's interests and concerns.

[1825] An "outlier" is a data point that is considered an anomaly in the collected data.

[1826] MODE FOR CARRYING OUT THE INVENTION

[1827] The present invention is a system that collects learning data for a generative AI by having users answer quiz-style tasks, and then trains an AI model based on this data. In this embodiment, the roles and specific operations of the user, terminal, and server are described.

[1828] User Authentication and Login

[1829] The user launches the smartphone app and logs in by entering their username and password on the login screen. This information is sent from the device to the server. The server references a database and verifies the authentication information. If authentication is successful, the server generates a login token and sends it to the device. The user is notified that the login was successful and is taken to the main screen.

[1830] Submitting a Quiz Task

[1831] After logging in, the server checks the user's usage history and generates a personalized quiz task. For example, it may ask a question such as "Which genre are you most interested in?" and provide the options "Movies," "Music," "Sports," and "News." The generated quiz task is sent to the terminal and notified to the user.

[1832] User responses and data submission

[1833] Users answer the quiz and press the submit button to send the answers from their device to the server. The server receives the answer data and removes inappropriate data and outliers. This clean data is added to the training dataset of the generative AI.

[1834] Data Processing and Validation

[1835] The server analyzes the collected response data and uses it as a learning dataset to train an AI model. This enables personalized content recommendations based on individual users' interests and preferences. The trained AI model is generated as a plugin and provided to companies in specific fields via API.

[1836] User Rewards and Feedback

[1837] Every time a user completes a quiz task, the server calculates the reward and adds it to the user's account. The terminal displays a notification to the user that points have been earned, encouraging them to participate in the next quiz task.

[1838] Hardware and software used

[1839] Hardware: Server, user's smartphone

[1840] Software: Flask (web application framework), SQLite (database), scikit-learn (machine learning library)

[1841] Specific examples

[1842] Consider the example of a prompt sentence in which a user answers a quiz on a smartphone app: "Which genre are you most interested in?"

[1843] Example prompt sentence:

[1844] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[1845] This allows users to enjoy content that interests them, and allows service providers to collect highly accurate data.

[1846] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1847] Program processing steps

[1848] Step 1:

[1849] The user launches the smartphone app and enters their username and password on the login screen. The entered authentication information is sent from the device to the server. The server receives this authentication information and authenticates the user by checking it against a database. If authentication is successful, the server generates a login token and sends it to the device. The device displays a notification to the user that login was successful and transitions to the main screen.

[1850] Input: Username, Password

[1851] Output: Login token, login success notification

[1852] Step 2:

[1853] The server checks the usage history of the logged-in user and generates a personalized quiz task based on that data. The content of the quiz task is, for example, a question such as "Which genre are you most interested in?" with multiple choices ("Movies," "Music," "Sports," "News"). The generated quiz task is sent to the terminal and notified to the user.

[1854] Input: User ID, usage history

[1855] Output: Quiz task

[1856] Step 3:

[1857] The user answers the quiz, selects one option, and presses the send button. This answer is sent from the device to the server.

[1858] Input: User's quiz answer

[1859] Output: Response data

[1860] Step 4:

[1861] The server receives the user's response data and performs analysis. Inappropriate data and outliers are removed, and the clean data is added to the training dataset of the generative AI.

[1862] Input: Response data

[1863] Output: Clean dataset

[1864] Step 5:

[1865] The server trains an AI model based on the clean data, which is then used to recommend personalized content. Once trained, the AI ​​model is generated as a plugin and made available to companies via API.

[1866] Input: Clean dataset

[1867] Output: Trained AI model, API plugin

[1868] Step 6:

[1869] When a user completes a quiz task, the server calculates the reward and adds it to the user's account. The device displays a notification to the user that points have been earned, encouraging them to participate in the quiz again.

[1870] Input: User ID, Quiz task completion information

[1871] Output: Reward points, notification

[1872] Example operation

[1873] Add specific examples of processing for each step.

[1874] Example of Step 1:

[1875] The user enters "alice@example.com" and "password123" into the smartphone app, the server authenticates, generates a login token, and returns it.

[1876] Example of Step 2:

[1877] The server checks the usage history of user "alice" and generates a quiz task asking "Which genre are you most interested in?" and provides the options "Movies," "Music," "Sports," and "News."

[1878] Example of Step 3:

[1879] The user selects "sports" and the answer data is sent to the server.

[1880] Example of Step 4:

[1881] The server analyzes the "Sports" responses, removes irrelevant data, and adds it to a clean dataset.

[1882] Example of Step 5:

[1883] The server uses the clean data collected to train an AI model and provides it to companies as a personalized content recommendation API.

[1884] Example of Step 6:

[1885] The server awards 10 points to user "alice" for completing the quiz task and displays a "Points Earned" notification.

[1886] Prompt Sentence Examples

[1887] "Take this quiz: What genre are you most interested in? Movies, music, sports, news?"

[1888] 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.

[1889] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state, by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. Specific embodiments of this system are described below.

[1890] System Overview

[1891] The system consists of the following main components:

[1892] 1. User authentication system

[1893] 2. Quiz Task Providing System

[1894] 3. Response data processing system

[1895] 4. Emotion Engine

[1896] 5. Learning Dataset Management System

[1897] 6. AI Model Training System

[1898] 7. Plugin Creation and Distribution System

[1899] 8. Reward Management System

[1900] User Authentication and Login

[1901] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1902] The terminal transmits the entered user authentication information to the server.

[1903] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1904] The device will notify the user that login was successful and transition to the main screen.

[1905] Submitting a Quiz Task

[1906] After logging in, the server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[1907] The server generates a quiz task such as "Which of the following products will be the best-selling in 2022?" and presents multiple options (e.g., smartphone, tablet, laptop, smartwatch).

[1908] The server sends the generated quiz task to the terminal.

[1909] The terminal notifies the user of the quiz task and displays a quiz screen.

[1910] User responses and data submission

[1911] The user selects one of the options displayed on the quiz screen and presses the submit button (e.g., selects smartphone).

[1912] The terminal transmits the selected answer to the server.

[1913] Data Processing and Sentiment Analysis

[1914] The server temporarily stores the received response data and starts data cleaning.

[1915] The server analyzes the received response data and removes inappropriate data and outliers.

[1916] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[1917] Learning dataset and AI model training

[1918] The server adds the clean data and emotion data to the AI's training dataset.

[1919] The server trains the AI ​​model based on the clean data and emotion data.

[1920] For example, quiz answers and emotional data collected from multiple users can be fed into an AI model to improve prediction accuracy.

[1921] Plugin generation and provision

[1922] The server generates a trained AI model as a generative AI plugin for a specific field.

[1923] The server provides the generated plugins to companies in specific fields or their own services via API.

[1924] Compensation management

[1925] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[1926] The terminal displays a notification to the user that points have been earned.

[1927] Specific examples

[1928] 1. User authentication and login details example

[1929] The user launches the app, enters "alice@example.com" and "password123" on the login screen, and clicks the login button.

[1930] The terminal sends this information to the server, which checks the information against a database.

[1931] If the authentication is successful, the server generates a token and returns it to the terminal.

[1932] The device will notify you that the login was successful and will transition to the main screen.

[1933] 2. Example of Quiz Task Submission Details

[1934] The server checks the usage history and emotional state of user "alice" and determines that it is time for her to watch an advertisement.

[1935] The emotion engine analyzes "Alice's" emotions in real time and determines, for example, that she is "relaxed."

[1936] The server generates a quiz task asking, "Which of the following products will be the best-selling in 2022?" and presents the following options: "smartphone," "tablet," "laptop," and "smartwatch."

[1937] The terminal displays the quiz and notifies the user.

[1938] 3. Detailed example of user responses and data submission

[1939] The user selects "smartphone" and presses the send button.

[1940] The terminal transmits this response data to the server.

[1941] The server analyzes the received response data and removes inappropriate data.

[1942] 4. Detailed examples of data processing and sentiment analysis

[1943] The server adds "smartphone" to the training dataset as clean data.

[1944] The emotion engine collects emotional data (e.g., "joy") from users' responses and adds it to the learning dataset.

[1945] 5. Detailed example of creating and providing a plugin

[1946] The server trains the AI ​​model based on the learning data and generates a "product trend prediction" plugin.

[1947] The server provides this plugin to companies and their own services via API.

[1948] 6. Detailed Example of User Rewards and Feedback

[1949] The server calculates a reward of 10 points for "alice" and adds it to her account.

[1950] The terminal displays a notification of the points earned, motivating the user to participate in the next quiz task.

[1951] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions to generate more effective learning data and improve the performance of the generation AI.

[1952] The processing flow will be explained below.

[1953] Step 1:

[1954] The terminal starts the application and displays the login screen. The user enters their username and password and presses the login button.

[1955] Step 2:

[1956] The terminal transmits the entered user authentication information to the server.

[1957] Step 3:

[1958] The server compares the received authentication information with the database, and if authentication is successful, it generates a login token and returns it to the terminal.

[1959] Step 4:

[1960] The device will notify the user that login was successful and transition to the main screen.

[1961] Step 5:

[1962] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz tasks based on that.

[1963] Step 6:

[1964] The emotion engine analyzes the user's emotions and adjusts the content and difficulty of the quiz task according to their state. For example, if the user is relaxed, a more difficult quiz is presented, and if the user is stressed, an easier quiz is presented.

[1965] Step 7:

[1966] The server generates a quiz task that asks the question, "Which of the following products will be the best-selling in 2022?" and provides several options (e.g., smartphone, tablet, laptop, smartwatch).

[1967] Step 8:

[1968] The server sends the generated quiz task to the terminal.

[1969] Step 9:

[1970] The terminal notifies the user of the quiz task and displays a quiz screen.

[1971] Step 10:

[1972] The user selects one of the options displayed on the quiz screen, for example, "smartphone," and presses the send button.

[1973] Step 11:

[1974] The terminal transmits the selected answer to the server.

[1975] Step 12:

[1976] The server temporarily stores the received response data and starts cleaning the data.

[1977] Step 13:

[1978] The server analyzes the received response data and removes inappropriate data and outliers.

[1979] Step 14:

[1980] The emotion engine collects emotional data (e.g., joy, surprise, annoyance) from users when they answer a quiz and adds it to the training dataset.

[1981] Step 15:

[1982] The server adds the clean data and emotion data to the AI's training dataset.

[1983] Step 16:

[1984] The server trains the AI ​​model based on the clean data and emotion data.

[1985] Step 17:

[1986] The server generates a trained AI model as a generative AI plugin for a specific field.

[1987] Step 18:

[1988] The server provides the generated plugins to companies in specific fields or their own services via API.

[1989] Step 19:

[1990] The server calculates the rewards (points or rewards) for users who complete the quiz tasks and adds them to the user's account.

[1991] Step 20:

[1992] The terminal displays a notification to the user that points have been earned.

[1993] Example 2

[1994] 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."

[1995] Conventional ad viewing formats require users to passively watch ads, resulting in poor user experience and limited advertising effectiveness. Furthermore, because interactive tasks that take into account the user's emotional state are not provided, it is difficult to effectively collect data based on the user's interests. Furthermore, there are issues with quality control of collected data and optimizing training datasets, which can lead to the inclusion of inappropriate data.

[1996] 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 inputting user authentication information and transmitting it to a device on the network; means for the device on the network to verify the authentication information and generate a login token if authenticated; and means for providing the authenticated user with a quiz task instead of viewing an advertisement. This enables the user's advertisement viewing experience to be replaced with an active quiz task. The server also includes means for the user to answer the quiz and transmit the answer to a device on the network; and means for the device on the network to receive the answer data, clean the data, and add it to a training dataset. This improves quality control of collected data and enables the generation of an appropriate training dataset. The server also includes means for training an artificial intelligence model based on the training dataset and generating the model as a plugin, means for providing the plugin to entities in a specific field via an API, and means for granting a reward each time a user completes a quiz task. This improves the performance of the AI ​​model and increases user motivation to participate.

[1997] A "networked device" is an electronic device such as a computer, smartphone, or tablet that can communicate over the Internet or other networks.

[1998] "Authentication Information" refers to the username, password, and other authentication means entered by a user when logging into a system.

[1999] A "login token" is a unique identifier or security token that is generated when a server successfully authenticates a user and is used to maintain the user's login session.

[2000] A "quiz task" is a question-type task in which the user can participate interactively, and in which the user selects the correct answer from multiple options.

[2001] "Data cleaning" is the process of automatically removing outliers and irrelevant data from received data, a process that aims to improve data quality.

[2002] A "training dataset" is a collection of data used to train an artificial intelligence model, including clean data and sentiment data.

[2003] An "artificial intelligence model" is a machine learning algorithm that is trained on a collected dataset and is a computer program to automatically perform a specific task.

[2004] A "plugin" is a software module that adds specific functionality or services to other systems or applications via an API.

[2005] "Entity" is a general term that refers to a legal entity, company, or organization that provides business or services in a particular field.

[2006] "API" is an abbreviation for Application Programming Interface, an interface for sharing functions between different software applications.

[2007] "Rewards" refer to incentives such as points or benefits that are given to users each time they complete a quiz task.

[2008] System Overview

[2009] The present invention is a system that combines a user-generated AI with an emotion engine that recognizes the user's emotional state by providing training data for the AI ​​through quiz-style tasks instead of watching advertisements. The system is implemented using devices (servers and terminals) on a network.

[2010] Hardware and software used

[2011] Hardware: Servers, devices (computers, smartphones, tablets)

[2012] Software: User authentication system, quiz task provision system, answer data processing system, emotion engine, learning dataset management system, AI model training system, reward management system

[2013] User Authentication and Login

[2014] The user starts the application using the device and the login screen is displayed. The user enters a username and password and sends them to the server. The server compares the received authentication information with the database, and if authentication is successful, generates a login token and sends it back to the device. The device displays a notification to the user that the login was successful and transitions to the main screen.

[2015] Submitting a Quiz Task

[2016] The server checks the user's usage history and emotional state based on the user ID. The emotion engine analyzes the user's emotions in real time and adjusts the difficulty and type of quiz task based on that. The server generates quiz tasks such as "Which of the following products was the best-selling in 2022?" and presents multiple options. The server sends the generated quiz task to the device, and the device notifies the user of the quiz task and displays the quiz screen.

[2017] User responses and data submission

[2018] The user selects one answer from the options displayed on the quiz screen and presses the send button. The device then sends the selected answer to the server.

[2019] Data Processing and Sentiment Analysis

[2020] The server temporarily stores the received answer data and starts data cleaning. The server analyzes the received answer data and removes inappropriate data and outliers. The emotion engine collects emotional data (e.g., joy, surprise, irritation) from the user when answering the quiz and adds it to the training dataset.

[2021] Learning dataset and AI model training

[2022] The server adds the cleaned response data and sentiment data to the AI ​​training dataset, which the server uses to train the AI ​​model and improve its prediction accuracy.

[2023] Plugin generation and provision

[2024] The server generates the trained AI model as a generative AI plugin for a specific domain, and provides the generated plugin to entities in the specific domain via an API.

[2025] Compensation management

[2026] The server calculates the reward (points or rewards) for the user who completed the quiz task and adds it to the user's account. The terminal displays a notification to the user that points have been earned.

[2027] Specific examples

[2028] 1. Example of user authentication:

[2029] The user launches the app, enters "alice@example.com" and "password123," and presses the login button. The device sends the information to the server, which checks the information in the database. If authentication is successful, the server generates a token and sends it back to the device. The device notifies the user that "Login was successful" and returns to the main screen.

[2030] 2. Example of a quiz task:

[2031] The server checks the usage history and emotional state of user "alice," which is "relaxed." The server generates a quiz task asking, "Which of the following products was the best-selling product in 2022?" and presents the options "smartphone," "tablet," "laptop," and "smartwatch." The device displays the quiz screen and notifies the user.

[2032] 3. Example prompt:

[2033] "Which of the following products will be the best-selling in 2022?"

[2034] "Predict what kind of ad would be best based on the emotional data of the user when they answer the quiz."

[2035] In this way, the present invention replaces the user's advertising viewing experience with a quiz task and further analyzes the user's emotions, thereby generating more effective training data and improving the performance of the generation AI.

[2036] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2037] Step 1:

[2038] A user starts an application using a terminal and a login screen is displayed. The user enters a username and password. The terminal sends the entered authentication information (username and password) to the server. The username and password are used as input data, and that data is sent to the server.

[2039] Step 2:

[2040] The server compares the received authentication information with the database and generates a login token if authentication is successful. It searches the database for matching authentication information and obtains the result of successful authentication as output. If authentication is successful, a login token is generated and the server sends that login token to the terminal. The output is the login token if successful. As a specific example, the server compares the username "alice@example.com" with the password "password123" and generates a login token if successful.

[2041] Step 3:

[2042] Based on the login token received by the terminal from the server, the terminal displays a notification of successful login to the user and transitions to the main screen. The login token received from the server is used as input. As output, the terminal displays a "login successful" notification to the user and transitions to the main screen.

[2043] Step 4:

[2044] The server checks the usage history and emotional state based on the user ID. As input data, it compares past usage history and real-time emotional state data based on the user ID. As output, it obtains the confirmed usage history and the emotion analysis results. For example, the emotion engine analyzes the emotion of user "alice" as "relaxed."

[2045] Step 5:

[2046] The server generates quiz tasks by adjusting the difficulty and type of the quiz task based on the emotional state. The appropriate quiz task is determined based on the input data of the emotion analysis results and the user's usage history. The generated quiz task is obtained as output. As a specific example, the server generates a quiz question such as, "Which of the following products was the best-selling product in 2022?"

[2047] Step 6:

[2048] The server sends the generated quiz task to the terminal, and the terminal notifies the user of the quiz task and displays the quiz screen. As input, the terminal receives the quiz task sent from the server. As output, the quiz screen is displayed to the user. The terminal displays the quiz screen and notifies the user to "answer the quiz."

[2049] Step 7:

[2050] The user selects one of the options displayed on the quiz screen and presses the send button. The answer data selected by the user is collected as input data. The selected answer data is obtained as output. For example, the user selects "smartphone" and presses the send button.

[2051] Step 8:

[2052] The device sends the selected answer data to the server. As input, it receives the answer data selected by the user. As output, it obtains the answer data to be sent to the server. The device sends the answer data "smartphone" to the server.

[2053] Step 9:

[2054] The server temporarily stores the received response data and begins data cleaning. Inappropriate data and outliers are removed. The received response data is used as input. Cleaned, clean data is obtained as output. The server stores the "smartphone" as clean data and removes inappropriate data.

[2055] Step 10:

[2056] The emotion engine collects emotional data when users answer quizzes and adds it to the training dataset. It uses real-time emotional data when users answer quizzes as input. It obtains the collected emotional data as output. For example, it collects the user's emotion when answering a quiz as "joy."

[2057] Step 11:

[2058] The server adds the cleaned response data and emotion data to the AI's training dataset. The cleaned data and emotion data are used as input. The output is an updated training dataset.

[2059] Step 12:

[2060] The server trains the AI ​​model based on the training dataset to improve prediction accuracy. The training dataset is used as input. The output is a trained AI model.

[2061] Step 13:

[2062] The server generates a trained AI model as a generative AI plugin for a specific field. The trained AI model is used as input. The generated plugin is obtained as output.

[2063] Step 14:

[2064] The server provides the generated plugin to a domain-specific entity via an API, uses the generated plugin as input, and sends the provided plugin to the domain-specific entity as output.

[2065] Step 15:

[2066] The server calculates a reward for a user who completes a quiz task and adds it to the user's account. As input, it uses the quiz task completion data. As output, the calculated reward is added to the user's account. The server calculates 10 points for "alice" and adds it to the account.

[2067] Step 16:

[2068] The terminal displays a notification to the user that a reward has been earned. As input, it receives the reward data sent from the server. As output, it displays a notification to the user that a reward has been earned. The terminal notifies the user that "points have been earned," motivating the user to participate in the next quiz task.

[2069] (Application example 2)

[2070] 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."

[2071] Conventional advertising viewing methods have made it difficult to effectively deliver ads to users and accurately measure ad performance. Furthermore, they have been unable to optimize ads based on the user's emotional state, leaving a lack of means to improve the user's advertising experience. The present invention aims to achieve more effective and personalized ad delivery by measuring advertising effectiveness through quiz-style tasks and analyzing the user's emotional state.

[2072] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[2073] In this invention, the server includes means for inputting user authentication information and transmitting it to the server, means for the server to verify the authentication information and generate a login token if authenticated, means for providing the authenticated user with a quiz task instead of watching an advertisement, means for the user to answer the quiz and transmit the answer to the server, means for the server to receive the answer data, remove inappropriate data, and add the answer to a training dataset, means for the server to analyze the user's emotional data when answering the quiz and add the data to the training dataset, means for training an AI model based on the training dataset and generating the model as a plugin, means for providing the plugin to companies in specific fields via an API, and means for awarding a reward each time a user completes a quiz task. This makes it possible to analyze the user's emotional state and provide more effective advertisement delivery and a personalized user experience.

[2074] "User authentication information" is information for identifying users and controlling access.

[2075] A "login token" is a temporary session identifier for an authenticated user to access a system.

[2076] A "quiz task" is a challenge presented to a user consisting of a series of questions and multiple-choice options.

[2077] "Answer data" is information about the answer selected by the user for the quiz task.

[2078] "Inappropriate data" is data that is not worthy or reliable for the system to include in the training dataset.

[2079] "Emotion data" is data that indicates the user's psychological state and emotions.

[2080] A "training dataset" is a set of clean data and sentiment data used to train an AI model.

[2081] An "AI model" is a software architecture that is trained to solve a specific problem based on machine learning algorithms.

[2082] A "plug-in" is a software component that extends or adds specific functionality.

[2083] "API" stands for Application Programming Interface, an interface for exchanging data and functions between different software systems.

[2084] "Rewards" are incentives such as points or benefits that users receive by completing quiz tasks.

[2085] The present invention is a system that combines a user-generated AI with learning data provided by the AI ​​through quiz-style tasks instead of watching advertisements, and an emotion engine that recognizes the user's emotional state. To realize this system, the following steps must be performed:

[2086] First, the device collects user authentication information and sends it to the server. The server verifies the authentication information, and if authentication is successful, generates a login token and returns it to the device. Once authentication is complete, the server provides the user with a quiz-style task instead of watching an advertisement. This quiz task is customized based on the user's usage history and real-time emotional state.

[2087] When a user answers a quiz, the device sends the answer data to the server. The server analyzes the received answer data, removes inappropriate data and outliers, and adds the clean data to the training dataset. At the same time, the server uses an emotion engine to analyze the user's emotional data when answering the quiz and adds that data to the training dataset.

[2088] The server then trains a generative AI model based on this learning dataset. The resulting AI model is then generated as a generative AI plugin for a specific field. This plugin is then provided to companies in that field or to their own services via API.

[2089] Through this process, the server will reward users each time they complete a quiz task. The collected clean data and emotional data will improve the accuracy of the AI ​​model, allowing companies to provide more effective advertising and services.

[2090] The hardware and software used are as follows:

[2091] Smartphone: The device through which the user accesses and answers the quiz tasks.

[2092] Python: A programming language used to develop each component of the system.

[2093] Emotion analysis model: An AI model for analyzing a user's emotional state.

[2094] Server: Validates authentication information, provides quiz tasks, processes answer and sentiment data, trains AI models, generates and provides plugins, and grants rewards.

[2095] Examples:

[2096] For example, imagine a user using an app. The user logs in to the app using a smartphone. After successfully logging in, the server generates a quiz question: "What was the best-selling product in 2022?" and displays it on the smartphone. The user selects "smartphone" and submits the answer. The server analyzes the answer data and uses an emotion engine to collect the user's emotional data at the time of answering as "joy." Finally, the user receives a reward of 10 points.

[2097] Example prompts to input to a generative AI model:

[2098] "Answer the following question: Which of the following products was your top-selling item in 2022?"

[2099] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2100] Step 1:

[2101] The terminal obtains the user authentication information and sends it to the server.

[2102] Enter your username and password.

[2103] Output: Sends authentication information to the server.

[2104] How it works: A user launches an application on their smartphone and enters their username and password into the login screen. The device then securely transmits this information to the server using the HTTPS protocol.

[2105] Step 2:

[2106] The server verifies the authentication information and generates a login token if authenticated.

[2107] Input: The username and password submitted.

[2108] Output: Authentication result and login token.

[2109] What happens: The server checks the user information stored in the database against the authentication information sent. If authentication is successful, the server generates a temporary login token and sends it back to the device.

[2110] Step 3:

[2111] The server provides the authenticated user with a quiz task in exchange for viewing an advertisement.

[2112] Input: User ID and login token, usage history, emotional state.

[2113] Output: A customized quiz task.

[2114] How it works: The server checks the usage history and real-time emotional state of the authenticated user and generates a quiz task based on that. For example, it provides a question like "What will be the best-selling product in 2022?" with options (smartphone, tablet, laptop, smartwatch).

[2115] Step 4:

[2116] The terminal presents the generated quiz task to the user, and the user answers the quiz.

[2117] Input: A customized quiz task.

[2118] Output: The user's answer.

[2119] Specific operation: The device notifies the user of the quiz task, displays the question and options, and the user selects one and presses the answer button.

[2120] Step 5:

[2121] The terminal transmits the user's response data to the server.

[2122] Input: The user's answer.

[2123] Output: Send the response data to the server.

[2124] Specific operation: When the user selects an answer to the quiz and presses the send button, the device sends the answer data to the server.

[2125] Step 6:

[2126] The server analyzes the received response data and removes inappropriate data.

[2127] Input: User response data.

[2128] Output: Clean data.

[2129] Specific operation: The server validates the received response data, detects and removes inappropriate data and outliers, and only reliable data is added to the training dataset.

[2130] Step 7:

[2131] The server uses an emotion engine to analyze the emotion data of the user when answering the quiz.

[2132] Input: User response data.

[2133] Output: User emotion data.

[2134] Specific operation: The server uses an emotion engine to analyze the user's facial expressions and voice patterns when answering the quiz, and identifies emotions such as joy, surprise, and irritation.

[2135] Step 8:

[2136] The server adds the clean data and the emotion data to the training dataset.

[2137] Input: Clean data, sentiment data.

[2138] Output: Added to the training dataset.

[2139] Specific operation: The server stores the clean data and analyzed emotion data in a learning dataset and uses it as training data for the generative AI model.

[2140] Step 9:

[2141] The server trains a generative AI model based on the learning dataset and generates the model as a plugin.

[2142] Input: The training dataset.

[2143] Output: The trained AI model plugin.

[2144] What it does: The server runs machine learning algorithms on a training dataset to train a generative AI model for use in a specific domain, and outputs the completed model in the form of a plugin.

[2145] Step 10:

[2146] The server provides the generated AI plugins to companies in specific fields via API.

[2147] Input: A trained AI model plugin.

[2148] Output: Provided via API.

[2149] How it works: The server provides trained AI model plugins to companies in specific fields via an API interface, which then use the plugin's functions in their own systems.

[2150] Step 11:

[2151] The server awards a reward each time a user completes a quiz task.

[2152] Input: User ID.

[2153] Output: Reward points.

[2154] How it works: The server monitors the user's quiz task completion status and adds a set reward point to the user's account every time the task is completed. This reward motivates the user.

[2155] 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.

[2156] 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.

[2157] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2158] 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.

[2159] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2160] 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.

[2161] 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).

[2162] 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.

[2163] 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."

[2164] 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.

[2165] 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).

[2166] 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.

[2167] 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.

[2168] 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.

[2169] 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.

[2170] 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.

[2171] 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.

[2172] 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.

[2173] 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.

[2174] 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 ...

Claims

1. means for inputting and transmitting user authentication information to a server; A means for the server to validate the credentials and, if authenticated, generate a login token; means for providing authenticated users with a quiz task in exchange for viewing an advertisement; A means for a user to answer a quiz and transmit the answer to a server; A means for the server to receive the response data, remove inappropriate data, and add the data to the training dataset; A means to train an AI model based on a training dataset and generate that model as a plugin; A way to provide plugins to companies in specific fields via API, a means for awarding a reward to a user each time the user completes a quiz task; A system including:

2. 2. The system according to claim 1, wherein the server further comprises means for checking the user's usage history.

3. The system of claim 1 , wherein the server further comprises means for automatically analyzing the answer data of the quiz task and removing impure outliers.

Images