System

Abstract

A system is provided.SOLUTION: A system comprising: means for creating a market that provides a generated AI model specialized in a specific field; means for storing the generated AI model in a database; means for an end user to access and retrieve the model and display detailed information; means for purchasing and paying for the generated AI model; means for tuning the generated AI model in response to a customization request; and means for monitoring usage of the generated AI model and performing necessary maintenance.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] While generative AI is beginning to be used in a wide range of fields, there is a problem in that general-purpose generative AI models cannot deliver the expected performance in specific fields. This makes it difficult to provide generative AI models with the required performance specialized for specific fields. Furthermore, when tuning or customizing a generative AI model is required, this work requires advanced expertise, making it a high hurdle for many users. Furthermore, effective operation and maintenance of the model often poses challenges in actual usage situations. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by building a marketplace that provides generative AI models specialized for specific fields and providing a system that users can easily access, search, purchase, and use. Specifically, it stores specialized generative AI models in a database and provides a means for users to view detailed information about the models and purchase them. It also includes a function to forward customization requests to experts and re-provide tuned generative AI models. Furthermore, by building a system that monitors the usage of generative AI models, notifies users when an abnormality is detected, and performs necessary maintenance, it is possible to realize the effective use of generative AI in specific fields.

[0006] A "generative AI model specialized for a specific field" is an AI model that is optimized for a specific industry or application and exhibits higher performance than general generative AI models.

[0007] "Market" means an online platform that enables users to search, purchase, and use domain-specific generative AI models.

[0008] "User" refers to an individual or company that uses a domain-specific generative AI model.

[0009] A "database" is a digital storage system for storing and managing domain-specific generative AI models.

[0010] "Detailed Information" refers to information such as characteristics, features, use cases, and performance metrics related to the Generative AI Model.

[0011] A "Customization Request" is a request submitted by a User to adjust or improve a Generative AI Model for a specific Use.

[0012] "Experts" refer to people with advanced knowledge and skills to tune and customize generative AI models.

[0013] "Tuning" is the process of adjusting a generative AI model to optimize its performance for a specific field or application.

[0014] "Monitoring" refers to the activity of overseeing the usage and performance of generative AI models to detect anomalies and problems.

[0015] "Maintenance" refers to the maintenance work required to fix problems that arise during the operation of a generative AI model and to keep it operating in optimal condition at all times.

[0016] An "anomaly" refers to behavior or errors that deviate from normal operation and occur during the use of a generative AI model.

[0017] "Notification" is a message that informs the user when an abnormality or problem occurs. [Brief explanation of the drawings]

[0018] [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

[0019] 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.

[0020] First, the terms used in the following description will be explained.

[0021] 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).

[0022] 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.

[0023] 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.

[0024] 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.

[0025] 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."

[0026] [First embodiment]

[0027] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0028] 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.

[0029] 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).

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] 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."

[0039] A system for providing generative AI models specialized for specific fields is configured as follows.

[0040] System Overview

[0041] This system builds a marketplace that provides generative AI models specialized in specific fields, and allows users to access the marketplace to search, purchase, and use models. The system mainly consists of the following components:

[0042] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[0043] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes databases, authentication systems, and payment systems.

[0044] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, and transaction history.

[0045] Explanation of program processing

[0046] User Registration and Authentication

[0047] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[0048] Terminal: Sends the entered information to the server as an HTTP request.

[0049] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[0050] Server: Returns a successful registration message to the user.

[0051] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[0052] Model Exploration and Selection

[0053] Users: Explore generative AI models optimized for specific domains (e.g., healthcare, finance, etc.).

[0054] Device: Sends the user's search query to the server.

[0055] Server: Retrieves a list of relevant generative AI models from the database and returns them to the user.

[0056] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[0057] Server: Retrieves detailed information from the database and sends it to the device.

[0058] Terminal: Display detailed information to the user.

[0059] Purchasing and using models

[0060] User: Select the model they wish to purchase and enter their payment information to complete the purchase.

[0061] Terminal: Sends the entered payment information to the server.

[0062] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[0063] Users: Download purchased models or access them through API.

[0064] Model tuning and customization

[0065] User: If a purchased model needs customization, the user sends a request to the server with their specific requirements.

[0066] Server: Receives requests and relays them to expert data scientists.

[0067] Server: Data scientists tune the model and generate a new model.

[0068] Server: When a new model is ready, it notifies the user and provides access.

[0069] Monitoring and Maintenance

[0070] Server: Monitors the usage of the generative AI model in real time.

[0071] Server: Collects and analyzes usage data and notifies users if anomalies are detected, such as when a model responds excessively slowly or when the error rate is high.

[0072] Server: Once an abnormality is confirmed, maintenance work will be carried out and users will be notified that the service is available again once the problem has been fixed.

[0073] Specific examples

[0074] 1. User Registration and Authentication Example

[0075] Researchers at medical institutions register an account to use the system for the first time.

[0076] Researchers log in and search for generative AI models to find AI models specialized for pathological diagnosis.

[0077] 2. Example of model purchase

[0078] Researchers check detailed information about the pathology diagnostic AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[0079] After purchase, researchers can download the model and begin using it in their actual research.

[0080] 3. Model Tuning Example

[0081] A researcher requests customization, requesting detailed tuning based on clinical data.

[0082] Specialized data scientists will handle the process and provide newly tuned models.

[0083] This invention makes it possible to safely and effectively provide generative AI models specialized for specific fields, enabling users to expect high effectiveness in those fields. In addition, the ease of customization and maintenance improves user convenience.

[0084] The processing flow will be explained below.

[0085] User Registration and Authentication

[0086] Step 1:

[0087] User: Visit the Market and open the account creation page.

[0088] Step 2:

[0089] Device: Enter the required information (email address, password, username, etc.).

[0090] Step 3:

[0091] Terminal: Sends the entered information to the server as an HTTP request.

[0092] Step 4:

[0093] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[0094] Step 5:

[0095] Server: Returns a successful registration message to the user.

[0096] Step 6:

[0097] User: Opens the login page and enters their email address and password.

[0098] Step 7:

[0099] Terminal: Sends the entered authentication information to the server.

[0100] Step 8:

[0101] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[0102] Step 9:

[0103] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[0104] Model Exploration and Selection

[0105] Step 1:

[0106] User: Select a specific sector within the market (e.g., healthcare, finance, etc.).

[0107] Step 2:

[0108] On your device: Enter your search query in the search bar within Market.

[0109] Step 3:

[0110] Device: Sends a search query to the server.

[0111] Step 4:

[0112] Server: Obtain a list of relevant generative AI models from the database.

[0113] Step 5:

[0114] Server: Returns the retrieved model list to the user.

[0115] Step 6:

[0116] User: Select the model of interest from the returned list of models.

[0117] Step 7:

[0118] User: Sends a request to view detailed information about a selected model.

[0119] Step 8:

[0120] Server: Retrieves detailed information from the database and sends it to the device.

[0121] Step 9:

[0122] Terminal: Display detailed information to the user.

[0123] Purchasing and using models

[0124] Step 1:

[0125] User: Clicks the purchase button and enters payment information.

[0126] Step 2:

[0127] Terminal: Sends purchase request and payment information to the server.

[0128] Step 3:

[0129] Server: Processes the payment, and if successful, sends the user a purchase confirmation message and grants them access to the model.

[0130] Step 4:

[0131] Users: Download purchased models or access them via API.

[0132] Model tuning and customization

[0133] Step 1:

[0134] Users: If you require customization of the purchased model, please describe your request in detail.

[0135] Step 2:

[0136] Device: Sends a customization request to the server.

[0137] Step 3:

[0138] Server: Forwards the request to an expert data scientist.

[0139] Step 4:

[0140] Server: Data scientists perform tuning and create new generative AI models.

[0141] Step 5:

[0142] Server: Saves the new generative AI model in the database and notifies the user.

[0143] Step 6:

[0144] User: Use the improved model with the new access information.

[0145] Monitoring and Maintenance

[0146] Step 1:

[0147] Server: Monitors the usage of the generative AI model in real time.

[0148] Step 2:

[0149] Server: Collects and analyzes usage data.

[0150] Step 3:

[0151] Server: If an anomaly is detected, a notification is sent to the user.

[0152] Step 4:

[0153] Users: Receive notifications and take action as needed.

[0154] Step 5:

[0155] Server: If a problem is identified, maintenance work is carried out to fix it.

[0156] Step 6:

[0157] Server: Once the fix is ​​complete, notify users that it is available again.

[0158] In this way, the system provides generative AI models specialized for specific fields and can be effectively operated to meet the diverse needs of users.

[0159] Example 1

[0160] 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."

[0161] Platforms that provide generative AI models specialized in specific fields are required to build an environment where users can efficiently search, purchase, and use generative AI models. However, many current systems lack sufficient functionality for user authentication, search query analysis, customization request response, usage monitoring, and anomaly detection, making it difficult to improve user convenience and reliability.

[0162] 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.

[0163] In this invention, the server includes means for building a platform that provides generative AI models specialized in specific fields, means for storing generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for generative AI models, means for adjusting generative AI models in response to customization requests, means for monitoring the usage status of generative AI models and performing necessary maintenance work, means for users to log in and input authentication information, and means for analyzing user search queries and suggesting appropriate generative AI models. This enables users to efficiently search for, purchase, customize, and use generative AI models.

[0164] "Specific fields" refer to areas with specific expertise or needs, such as medicine, finance, or education.

[0165] "Generative AI models" refer to artificial intelligence models that generate new data and content using techniques such as generative adversarial networks (GANs) and natural language generation (NLG).

[0166] "Platform" refers to the infrastructure that provides an online environment for users to search for, purchase, and use generative AI models.

[0167] "Database" refers to a system for systematically storing and managing generative AI models, related user information, transaction history, etc.

[0168] "End User" refers to the ultimate user who utilizes the generative AI model to provide a specific application or service.

[0169] A "search query" refers to a keyword or phrase entered by a user when searching for a generative AI model.

[0170] "Customization Request" means a request submitted by a User to adjust or refine a Generative AI Model based on their specific needs.

[0171] "Maintenance work" refers to regular checks and maintenance work to ensure that generative AI models function properly.

[0172] "Login" refers to the authentication procedure required for a user to access a system.

[0173] "Authentication Information" refers to information such as email address and password that a User provides to prove access to a System.

[0174] "Search query analysis" refers to the process of analyzing a search query entered by a user and proposing the optimal generative AI model.

[0175] The system of the present invention builds a platform that provides generative AI models specialized in specific fields, and allows end users to search, purchase, customize, and use generative AI models using this platform. The main components of the system are as follows:

[0176] 1. User Device

[0177] A device that users access and operate. It can be a PC, smartphone, tablet, etc. Users access the system through their device to register an account, log in, search for models, view detailed information, make purchases, and request customization.

[0178] 2. Server

[0179] The server is the central component that manages and operates the entire system. It stores generative AI models in a database and includes an authentication system for user authentication, a search query analysis system, a payment system, and a system for responding to customization requests. It also provides API endpoints for users to access. The server communicates with the database and returns appropriate information based on the user's request.

[0180] 3. Database

[0181] This is a system for storing data such as generative AI models, user information, and transaction history. This database uses database systems such as PostgreSQL and Elasticsearch, particularly to speed up searches and maintain data consistency.

[0182] For example, a user accesses a market website and enters the required information (email address, password, username, etc.) on the account registration page. The device sends this information to the server, which validates it and stores it in the database. When the user enters their authentication information on the login page, the server authenticates the user and, if successful, allows them to access the market.

[0183] Next, the user searches for generative AI models suitable for a specific field (e.g., medicine or finance). The device sends the search query to the server, which retrieves and returns a list of relevant models from the database. The user selects the model of interest and requests that detailed information be displayed. The server retrieves the detailed information from the database and sends it to the device, allowing the user to purchase or customize the model.

[0184] Furthermore, if a user requests customization, the request is forwarded to the server and handled by a specialized data scientist. Once customization is complete, a new model is generated and provided to the user. The server also monitors the usage of the generated AI model in real time and takes appropriate action if an abnormality is detected.

[0185] Prompt Sentence Examples

[0186] 1. Researchers at medical institutions must register an account to use the system for the first time.

[0187] 2. Purchase and download the AI ​​model to be used for pathology diagnosis.

[0188] 3. Customize the pathology diagnostic AI model you purchased based on clinical data.

[0189] In this way, the system effectively provides domain-specific generative AI models, enhancing user convenience.

[0190] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0191] Step 1: User Registration

[0192] Users: Visit the Market registration page and enter your name, email address, and password.

[0193] Input: Name, email address, and password entered by the user into the form.

[0194] Output: Registration successful message.

[0195] Terminal: Sends user input information to the server in JSON format.

[0196] Input: Information entered by the user.

[0197] Output: HTTP POST request in JSON format.

[0198] Server: Validate the received information using a validation library (e.g. Joi) and, if there are no problems, save the user information to the PostgreSQL database.

[0199] Input: User information in JSON format.

[0200] Output: The new user information is saved in the database.

[0201] Server: Returns a successful registration message to the user in JSON format.

[0202] Input: User information saved successfully.

[0203] Output: Registration successful message.

[0204] Step 2: User Login

[0205] User: Access the login page and enter the registered email address and password.

[0206] Enter your email address and password.

[0207] Output: Login request.

[0208] Terminal: Sends the user's authentication information to the server in a JSON-formatted HTTP POST request.

[0209] Input: The credentials entered by the user.

[0210] Output: HTTP POST request in JSON format.

[0211] Server: Validates the received authentication information using a password authentication system (e.g. bcrypt) and issues a JWT token if it matches.

[0212] Input: User credentials.

[0213] Output: JWT token.

[0214] Server: Sends the JWT token to the user terminal.

[0215] Input: JWT token.

[0216] Output: User authentication success message and JWT token.

[0217] Step 3: Model Search

[0218] Users: Find generative AI models optimized for specific domains through search queries.

[0219] Input: Search query.

[0220] Output: The search request.

[0221] Terminal: Sends the user's query to the server as an HTTP GET request in JSON format.

[0222] Input: Search query.

[0223] Output: HTTP GET request in JSON format.

[0224] Server: Uses a search engine such as Elasticsearch to retrieve the relevant generative AI model from the database.

[0225] Input: Search query.

[0226] Output: A list of generative AI models.

[0227] Server: Returns search results to the user's device.

[0228] Input: A list of generative AI models.

[0229] Output: Search results.

[0230] Step 4: View model details

[0231] User: Select the model they are interested in and request more information.

[0232] Input: Model ID.

[0233] Output: Model details request.

[0234] Device: Send an HTTP GET request in JSON format containing the model ID to the server.

[0235] Input: Model ID.

[0236] Output: HTTP GET request in JSON format.

[0237] Server: Obtains detailed information about the selected generative AI model from the database and sends it to the user's device.

[0238] Input: Model ID.

[0239] Output: Detailed information about the model.

[0240] Terminal: Displays the received detailed information to the user.

[0241] Input: Model details.

[0242] Output: The detailed information displayed to the user.

[0243] Step 5: Purchase a model

[0244] User: Selects the model they want to purchase and enters their payment information to begin the checkout process.

[0245] Input: Payment information.

[0246] Output: Purchase request.

[0247] Terminal: Sends the user's payment information to the server via an HTTPS POST request in JSON format.

[0248] Input: Payment information.

[0249] Output: HTTPS POST request in JSON format.

[0250] Server: Process the payment using the Stripe API, and if successful, record the purchase confirmation message and model access rights in the database.

[0251] Input: Payment information.

[0252] Output: Purchase successful message and updated access rights.

[0253] Server: Sends a purchase confirmation message and access right information to the user terminal.

[0254] Input: Purchase success message and access rights information.

[0255] Output: Notification to user terminal.

[0256] Step 6: Use the model

[0257] Users: Access or download purchased models via API.

[0258] Input: Access request.

[0259] Output: Model use.

[0260] Server: Receives requests to access a model, performs appropriate authentication, and returns the model file or issues an API key.

[0261] Input: Access request.

[0262] Output: Model file or API key.

[0263] Terminal: Provide the received model file or API key to the user.

[0264] Input: Model file or API key.

[0265] Output: The model is made available to the user.

[0266] (Application example 1)

[0267] 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."

[0268] Modern factories require efficient and rapid optimization of manufacturing processes. It is difficult to easily obtain and use generative AI models suited to specific manufacturing environments. Furthermore, customization of generative AI models and effective monitoring after their deployment are required. To address these challenges, an effective means is needed to easily provide AI models for optimizing the operation of factory robots, and to customize and maintain those models.

[0269] 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.

[0270] In this invention, the server includes means for building a marketplace that provides generative AI models specialized in specific fields, means for storing the generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for the generative AI models, means for tuning the generative AI models in response to customization requests, means for monitoring the usage status of the generative AI models and performing necessary maintenance, means for end users to select, download, or integrate generative AI models to be applied to factory robots, and means for optimizing the operation of the factory robots for specific manufacturing processes. This enables end users to easily obtain generative AI models suitable for their manufacturing environment and effectively apply and operate them.

[0271] A "generative AI model" is an algorithmic model of artificial intelligence that is generated for a specific field or application, and exhibits performance optimized for that field.

[0272] "Market" means the online platform where end users can find, purchase, customize, download, and integrate Generative AI Models.

[0273] A "database" is an information recording system for managing and storing generative AI models, user information, transaction history, etc.

[0274] "End users" are the final users of generative AI models, such as factory workers and managers.

[0275] "Customization Request" means a request by an End User to tailor a Generative AI Model based on their specific needs and requirements.

[0276] "Tuning" refers to the process of adjusting a generative AI model to optimize it for the end user's specific use and environment.

[0277] "Monitoring" refers to the act of monitoring the usage and performance of generative AI models in real time.

[0278] "Maintenance" refers to maintenance work to maintain the performance and reliability of the generative AI model, including fixing bugs and making improvements.

[0279] A "factory robot" is a mechanical device used to automate specific tasks in manufacturing sites, and its operations are controlled by a program.

[0280] "Manufacturing process" refers to the series of operations or steps that transform raw materials into a finished product, including welding, assembly, quality inspection, etc.

[0281] "Integration" refers to the act of end users applying purchased generative AI models to their own systems or machinery (e.g., factory robots) to make them work together.

[0282] This invention relates to a system that enables end users to easily search, purchase, customize, download, and integrate generative AI models suited to specific manufacturing processes. This system is primarily composed of the following hardware and software:

[0283] Hardware and Software Use Cases

[0284] Hardware: A factory robot (e.g., a generic robotic arm) and the smartphone or tablet (iOS or Android device) that controls it.

[0285] software:

[0286] Robot control software (e.g., Robot Operating System (ROS)).

[0287] A cloud service (e.g., AWS or Google Cloud) as a model management server.

[0288] The application for operation on smart devices will be developed using React Native.

[0289] Program processing explanation

[0290] In this invention, the server, terminal, and user work together to search, select, purchase, customize, and use generative AI models. The specific processing steps are as follows:

[0291] 1. User Registration and Authentication

[0292] Users open the app on their smartphone and enter the required information on the account creation page.

[0293] The terminal transmits the input information to the server as an HTTP request.

[0294] The server validates the received information and, if there are no problems, saves the new user information in the database and returns a message to the user confirming successful registration.

[0295] 2. Model Search and Selection

[0296] Users enter a search query to explore generative AI models specific to their manufacturing process.

[0297] The terminal transmits the user's search query to the server.

[0298] The server retrieves a list of relevant generative AI models from the database and returns it to the user.

[0299] The user selects a model of interest from the returned list and requests that its detailed information be displayed.

[0300] The server retrieves the detailed information from the database and sends it to the terminal.

[0301] The terminal displays detailed information to the user.

[0302] 3. Purchasing and Using the Model

[0303] Users select the generative AI model they want to purchase and enter their payment information.

[0304] The terminal transmits the entered payment information to the server.

[0305] The server processes the payment and, if successful, gives the user a purchase confirmation message and access to the model.

[0306] Users can download the purchased model or integrate it into their robot control system via API.

[0307] 4. Tuning and customizing the model

[0308] When a user desires tuning specialized for a specific production line, the user writes a customization request and sends it to the server.

[0309] The server receives the request and relays it to the data scientist.

[0310] The server will notify the user and grant access as soon as a new tuned model is available.

[0311] 5. Monitoring and Maintenance

[0312] The server monitors the usage of the generative AI model in real time.

[0313] The server collects and analyzes usage data and notifies the user if an anomaly is detected, such as a robot slowing down or an increased error rate.

[0314] The server will perform maintenance when an abnormality is detected and notify users that the service is available again once the problem has been fixed.

[0315] Specific examples

[0316] For example, if an end user wants to apply an AI model to a welding process in a factory, they might enter a prompt like this:

[0317] "Optimize the welding seam angle and speed used by this robotic arm."

[0318] This allows the generative AI model to optimize the parameters of the welding process, reducing error rates and improving production speed.

[0319] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0320] Step 1: User Registration and Authentication

[0321] User: Opens the smartphone app and enters the required information (email address, password, factory ID, etc.) on the account creation page.

[0322] Terminal: Sends the entered information to the server as an HTTP request.

[0323] Server: Validates the received information and, if there are no problems, saves the new user information to the database. If the save is successful, generates a message indicating successful registration and returns it to the terminal. Examples of validation include detecting invalid email addresses and passwords that are too short.

[0324] Step 2: Log in

[0325] User: Enters email address and password on the login page and clicks the login button.

[0326] Terminal: Sends the entered information to the server as an HTTP request.

[0327] Server: Searches for the corresponding user information in the database and compares it with the authentication information. If it matches, it generates a login success message and session information and sends them to the terminal.

[0328] Step 3: Model exploration

[0329] User: Enter a specific manufacturing process (e.g., welding, assembly) into the app's search bar and press the search button.

[0330] Terminal: Sends the search query as an HTTP request to the server.

[0331] Server: Searches the database for a list of relevant generative AI models, generates a list of relevant models, and sends it to the device. Specifically, it filters the models by tags and keywords related to the manufacturing process.

[0332] Step 4: Model selection and detailed display

[0333] User: Selects the generative AI model of interest from the returned list and requests more information about it.

[0334] Device: Send the model ID to the server as an HTTP request.

[0335] Server: Retrieves detailed information about the selected model from the database, generates detailed information, and sends it to the terminal. The detailed information includes the model's scope of application, expected results, and usage methods.

[0336] Step 5: Purchase and pay for the model

[0337] User: Select the generative AI model they want to purchase and enter their payment information (credit card number, security code, etc.).

[0338] Terminal: Sends payment information to the server as an HTTP request.

[0339] Server: Works with the payment processing system to perform payment, and if successful, generates a purchase confirmation message and access rights for the model and sends them to the terminal. If unsuccessful, generates an error message and sends it to the terminal.

[0340] Step 6: Download and integrate the model

[0341] User: Presses a button to download the purchased generative AI model.

[0342] Device: Sends a model download request to the server.

[0343] Server: Retrieves the corresponding model file from the database, generates a download link, and sends it to the device.

[0344] Users: Click the download link and integrate the model file into their robot control system. Specifically, they load the model using software such as ROS and set parameters to optimize the robot's behavior.

[0345] Step 7: Tune and customize the model

[0346] User: If a user wants specialized tuning for a specific production line, he or she enters a customization request and sends it to the server.

[0347] Terminal: Sends the customization request to the server as an HTTP request.

[0348] Server: Receives requests and forwards them to data scientists, who tune the model based on the requests and generate new versions of the model.

[0349] Server: Saves the new tuned model to the database and notifies the user of the update.

[0350] Step 8: Monitoring and Maintenance

[0351] Server: Collects logging and performance metrics to monitor the usage of generative AI models in real time.

[0352] Server: If abnormal behavior (e.g., delayed response, increased error rate) is detected, the server identifies the anomaly based on an anomaly detection algorithm and notifies the user. Statistical anomaly detection techniques and machine learning algorithms are used for anomaly detection.

[0353] Server: If an anomaly is detected, the maintenance team is notified and the model is available again once the fix is ​​complete. Continuous monitoring ensures the reliability and effectiveness of the model.

[0354] 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.

[0355] The system for providing generative AI models specialized for specific fields is configured as follows: This system also incorporates an emotion engine that recognizes user emotions.

[0356] System Overview

[0357] This system creates a marketplace that provides generative AI models specialized in specific fields, allowing users to access the marketplace and search, purchase, and use models. It also has the ability to recognize user emotions and suggest models and interact with them based on those emotions.

[0358] The system mainly consists of the following components:

[0359] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[0360] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes a database, authentication system, payment system, emotion engine, etc.

[0361] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, transaction history, and emotional data.

[0362] 4. Emotion Engine: A component for recognizing and analyzing user emotions.

[0363] Explanation of program processing

[0364] User Registration and Authentication

[0365] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[0366] Terminal: Sends the entered information to the server as an HTTP request.

[0367] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[0368] Server: Returns a successful registration message to the user.

[0369] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[0370] Model Exploration and Selection

[0371] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[0372] Device: Sends the user's search query to the server.

[0373] Server: Retrieves a list of applicable generative AI models from the database and returns it to the user. The emotion engine recognizes the user's emotions and prioritizes models that evoke positive emotional responses.

[0374] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[0375] Server: Retrieves detailed information from the database and sends it to the device.

[0376] Terminal: Display detailed information to the user.

[0377] Purchasing and using models

[0378] User: Clicks the purchase button and enters payment information.

[0379] Terminal: Sends purchase request and payment information to the server.

[0380] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[0381] Users: Download purchased models or access them via API.

[0382] Model tuning and customization

[0383] Users: If you require customization of the purchased model, please describe your request in detail.

[0384] Device: Sends a customization request to the server.

[0385] Server: Forwards the request to an expert data scientist.

[0386] Server: Data scientists perform tuning and create new generative AI models.

[0387] Server: Saves the new generative AI model in the database and notifies the user.

[0388] User: Use the improved model with the new access information.

[0389] Monitoring and Maintenance

[0390] Server: Monitors the usage of the generative AI model in real time.

[0391] Server: Collects and analyzes usage data.

[0392] Server: If an anomaly is detected, a notification is sent to the user.

[0393] Users: Receive notifications and take action as needed.

[0394] Server: If a problem is identified, maintenance work is carried out to fix it.

[0395] Server: Once the fix is ​​complete, notify users that it is available again.

[0396] Incorporating an emotion engine

[0397] Server: Builds an emotion engine and collects emotion data in real time from users' voices, text, facial expressions, etc.

[0398] Server: The emotion engine analyzes the collected data and determines the emotional state.

[0399] Server: Based on the acquired emotion data, the server presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, the server prioritizes models that are effective in reducing stress.

[0400] Server: The emotion engine periodically analyzes the emotion data and suggests or customizes models as needed.

[0401] User: Review the sentiment-based recommendation model and make a purchase or use it.

[0402] Specific examples

[0403] 1. User Registration and Authentication Example

[0404] A pharmaceutical researcher from a healthcare company registers an account to use the system for the first time.

[0405] Researchers log in and search generative AI models to find AI models specialized for specific diseases.

[0406] 2. Example of model purchase

[0407] Researchers check detailed information about the disease-specific AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[0408] After purchase, researchers can download the model and begin using it in their research.

[0409] 3. Model Tuning Example

[0410] A researcher requests customization, requesting detailed tuning based on clinical data.

[0411] Specialized data scientists will handle the process and provide newly tuned models.

[0412] 4. Example of an Emotion Engine

[0413] As researchers use the system, the emotion engine analyzes their context and recognizes when they are feeling stressed.

[0414] Based on the emotion engine, we present effective guidelines for stress reduction and propose a model to improve researchers' work efficiency.

[0415] This invention enables the safe and effective provision of generative AI models specialized for specific fields. In addition, by incorporating an emotion engine, it is possible to provide more personalized services according to the user's emotional state.

[0416] The processing flow will be explained below.

[0417] User Registration and Authentication

[0418] Step 1:

[0419] User: Visit the Market and open the account creation page.

[0420] Step 2:

[0421] Device: Enter the required information (email address, password, username, etc.).

[0422] Step 3:

[0423] Terminal: Sends the entered information to the server as an HTTP request.

[0424] Step 4:

[0425] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[0426] Step 5:

[0427] Server: Returns a successful registration message to the user.

[0428] Step 6:

[0429] User: Opens the login page and enters their email address and password.

[0430] Step 7:

[0431] Terminal: Sends the entered authentication information to the server.

[0432] Step 8:

[0433] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[0434] Step 9:

[0435] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[0436] Model Exploration and Selection

[0437] Step 1:

[0438] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[0439] Step 2:

[0440] On your device: Enter your search query in the search bar within Market.

[0441] Step 3:

[0442] Device: Sends a search query to the server.

[0443] Step 4:

[0444] Server: Obtain a list of relevant generative AI models from the database.

[0445] Step 5:

[0446] Server: Returns the acquired model list to the user device.

[0447] Step 6:

[0448] User: Select the model of interest from the returned list of models.

[0449] Step 7:

[0450] User: Sends a request to view detailed information about a selected model.

[0451] Step 8:

[0452] Server: Retrieves detailed information from the database and sends it to the device.

[0453] Step 9:

[0454] Terminal: Display detailed information to the user.

[0455] Step 10:

[0456] Emotion engine: Analyzes the user's facial expressions, voice, text, etc. to obtain emotional data.

[0457] Step 11:

[0458] Server: Based on the emotion data, add the generative AI model that is suitable for the user to the suggestion list.

[0459] Purchasing and using models

[0460] Step 1:

[0461] User: Clicks the purchase button and enters payment information.

[0462] Step 2:

[0463] Terminal: Sends purchase request and payment information to the server.

[0464] Step 3:

[0465] Server: Processes the payment, and if successful, sends the user a purchase confirmation message and grants them access to the model.

[0466] Step 4:

[0467] Users: Download purchased models or access them through the API.

[0468] Model tuning and customization

[0469] Step 1:

[0470] Users: If you need customization for the model you purchased, please submit your specific request.

[0471] Step 2:

[0472] Terminal: Forwards the customization request to the server.

[0473] Step 3:

[0474] Server: Forwards the request to an expert data scientist.

[0475] Step 4:

[0476] Server: Data scientists tune the model and generate a new generative AI model.

[0477] Step 5:

[0478] Server: Saves the newly tuned generative AI model in the database and notifies the user.

[0479] Step 6:

[0480] User: Use the improved model with the new access information.

[0481] Monitoring and Maintenance

[0482] Step 1:

[0483] Server: Monitors the usage of the generative AI model in real time.

[0484] Step 2:

[0485] Server: Collects and analyzes usage data.

[0486] Step 3:

[0487] Server: If an anomaly is detected, a notification is sent to the user.

[0488] Step 4:

[0489] Users: Receive notifications and take action as needed.

[0490] Step 5:

[0491] Server: If a problem is identified, maintenance work is carried out to fix it.

[0492] Step 6:

[0493] Server: Once the fix is ​​complete, notify users that it is available again.

[0494] Incorporating an emotion engine

[0495] Step 1:

[0496] Server: Builds an emotion engine and collects emotion data in real time from users' voices, text, facial expressions, etc.

[0497] Step 2:

[0498] Server: The emotion engine analyzes the collected data and determines the emotional state.

[0499] Step 3:

[0500] Server: Based on the acquired emotion data, presents the user with a list of optimal generative AI models.

[0501] Step 4:

[0502] Server: Adjusts suggestions and interactions to the user based on emotion data. For example, if the user is feeling stressed, it makes suggestions to help them relax.

[0503] Specific examples

[0504] User registration and authentication examples

[0505] A researcher opens the account registration page.

[0506] The researcher enters the required information and completes the registration.

[0507] The researcher logs in using their email address and password.

[0508] Specific examples of model purchases

[0509] Researchers select a disease-specific AI model and view detailed information.

[0510] The researcher decides to purchase and enters payment information.

[0511] Researchers can download the model and use it in their research.

[0512] Example of model tuning

[0513] The researcher details and submits the customization request.

[0514] The server forwards the request to an expert data scientist.

[0515] Data scientists tune the model and provide a new model.

[0516] Examples of emotion engines

[0517] While researchers are using the system, an emotion engine detects their stress levels.

[0518] The server proposes a model suitable for stress reduction.

[0519] Researchers will follow the suggestions and use the model to improve operational efficiency.

[0520] This invention makes it possible to provide generative AI models specialized for specific fields while providing individual services according to the user's emotional state. The incorporation of an emotion engine is expected to improve the user experience and lead to more effective use.

[0521] Example 2

[0522] 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."

[0523] In recent years, the use of generative AI models has expanded in various fields, but there is a lack of systems that efficiently provide generative AI models that can meet specialized requirements.In addition, the lack of models that adapt to the user's emotional state often leads to a decrease in user satisfaction and hinders the effective use of the models.

[0524] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for building a market that provides generative AI models specialized in a specific field, a means for storing the generative AI models in a database, and a means for analyzing user emotion data and proposing an optimal generative AI model based on the analyzed data. This makes it possible to efficiently provide specialized generative AI models and propose and use models that are adapted to the user's emotional state.

[0525] "Market" is an online platform for providing domain-specific generative AI models.

[0526] A "generative AI model" is an artificial intelligence model that is generated based on data from a specific field and automates or assists with various tasks in that field.

[0527] A "database" is a system for storing generative AI models, user information, emotional data, etc.

[0528] "End User" means an individual or legal entity that uses the System to search for, purchase, and use Generative AI Models.

[0529] "Search Means" means means that provide the ability for an End User to search for a particular Generative AI Model.

[0530] "Means for displaying detailed information" refers to means for providing a function for end users to display detailed information (e.g., performance indicators, price, usage methods, etc.) about the generated AI model in which they are interested.

[0531] "Means for purchasing and payment" refers to means that provide the functionality for processing payments when end users purchase generative AI models.

[0532] A "customization request" is a request by an end user for additional functionality or tuning of a generative AI model that they have purchased.

[0533] "Means for tuning" means means for providing the ability to adjust or improve a generative AI model in response to end-user customization requests.

[0534] "Monitoring means" refers to a means for monitoring the usage of the generated AI model in real time and providing the functionality to detect anomalies and collect data.

[0535] "Maintenance measures" are measures for carrying out corrective work or maintenance to resolve abnormalities or problems detected through monitoring.

[0536] "Emotional data" is data that indicates the user's emotional state and is collected from voice, text, facial expressions, etc.

[0537] The "means for analyzing emotional data" is a means for analyzing collected emotional data and providing a function for determining the emotional state of the user.

[0538] The "emotion engine" is a system component that proposes and adjusts models based on the user's emotional state.

[0539] The present invention relates to a system that provides generative AI models specialized in specific fields. The system allows users to access, search, purchase, and use specialized generative AI models, and further has the function of analyzing user emotion data and proposing optimal generative AI models based on the data.

[0540] System Overview

[0541] The system consists of the following components:

[0542] 1. User terminal: A device that a user accesses and operates. Examples include PCs, smartphones, tablets, etc.

[0543] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. The server includes a database, authentication system, payment system, emotion engine, etc.

[0544] 3. Database: Stores data such as domain-specific generative AI models, user information, transaction history, and sentiment data.

[0545] 4. Emotion Engine: A component for recognizing and analyzing user emotions.

[0546] User Registration and Authentication

[0547] When a user accesses the market, they first enter the required information on the account creation page. The device sends this information to the server as an HTTP request. The server validates the received information and, if there are no problems, saves the new user information in the database. The user then enters their authentication information on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[0548] Model Exploration and Selection

[0549] When a user searches the market for a generative AI model optimized for a specific field, the device sends a search query to the server. The server retrieves a list of relevant generative AI models from the database and returns it to the user. The emotion engine recognizes the user's emotions and prioritizes models that evoke positive emotional reactions. The user selects a model of interest from the returned list of models and requests that its detailed information be displayed on the device. The server retrieves the detailed information from the database and sends it to the device. The device displays this information to the user.

[0550] Purchasing and using models

[0551] When a user purchases a model, they click the purchase button and enter their payment information. The device sends this request and payment information to the server. The server processes the payment through an external payment gateway, and if successful, issues the user a purchase confirmation message and access rights to the model. The user can then download the purchased model or access it via API.

[0552] Model tuning and customization

[0553] If a user wants to customize a model, they send a detailed request from their device to the server. The server then forwards the request to a specialized data scientist. Once the data scientist performs tuning and a new generative AI model is created, the server stores it in a database and notifies the user. The user can then use the improved model with the new access information.

[0554] Monitoring and Maintenance

[0555] The server monitors the usage of the generative AI model in real time, collecting and analyzing usage data. If an abnormality is detected, a notification is sent to the user. The user receives the notification and can take action as necessary. If a problem is confirmed, the server performs maintenance work, and once the problem is fixed, the server notifies the user that the model is available again.

[0556] Incorporating an emotion engine

[0557] The emotion engine allows the server to collect and analyze emotional data from the user's voice, text, facial expressions, etc. in real time. Based on the acquired emotional data, the server presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, models that are effective in reducing stress will be prioritized. The emotion engine regularly analyzes the emotional data and suggests or customizes models as needed.

[0558] Specific examples

[0559] 1. User Registration and Authentication Example

[0560] A pharmaceutical researcher from a healthcare company registers an account to use the system for the first time.

[0561] Researchers log in and search generative AI models to find AI models specialized for specific diseases.

[0562] 2. Example of model purchase

[0563] Researchers check detailed information about the disease-specific AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[0564] After purchase, researchers can download the model and begin using it in their research.

[0565] 3. Model Tuning Example

[0566] A researcher requests customization, requesting detailed tuning based on clinical data.

[0567] Specialized data scientists will handle the process and provide newly tuned models.

[0568] 4. Example of an Emotion Engine

[0569] As researchers use the system, the emotion engine analyzes their context and recognizes when they are feeling stressed.

[0570] Based on the emotion engine, we present effective guidelines for stress reduction and propose a model to improve researchers' work efficiency.

[0571] As a result, the present invention can safely and effectively provide generative AI models specialized for specific fields, while also providing personalized services that correspond to the user's emotional state.

[0572] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0573] User Registration and Authentication

[0574] Step 1:

[0575] The user accesses the market and enters the required information (email address, password, username, etc.) on the account creation page. The entered information is displayed on the user's device, and once confirmation is complete, the user presses the send button.

[0576] Step 2:

[0577] The device sends the entered information to the server as an HTTP request, which contains the entered information and is formatted in JSON.

[0578] Step 3:

[0579] The server validates the information it receives. For example, it checks the format of the email address or the strength of the password. This validation is performed using regular expressions or rule-based checks. If the input is found to be valid, it proceeds.

[0580] Step 4:

[0581] The server stores the successfully validated information in the database. The password is hashed and stored securely along with the user information. The storage process uses an SQL query.

[0582] Step 5:

[0583] The server returns a message to the user confirming successful registration. The message is structured in JSON format and sent to the terminal for display.

[0584] Step 6:

[0585] The user again enters their authentication information (email address, password) on the login page. The entered information is again sent by the terminal to the server as an HTTP request.

[0586] Step 7:

[0587] The server checks the received authentication information against its database, specifically to ensure that the hashed password matches. If the check is successful, it issues a session token to the user and returns a response containing that session token.

[0588] Step 8:

[0589] Users can access the market using a session token, which is stored in a browser cookie and used to simplify authentication in future visits.

[0590] Model Exploration and Selection

[0591] Step 1:

[0592] Users enter keywords in the Market's search bar to find generative AI models optimized for a specific field (e.g., healthcare, finance, etc.).

[0593] Step 2:

[0594] The device sends the entered search query to the server, which includes the keywords entered by the user.

[0595] Step 3:

[0596] The server retrieves a list of relevant generative AI models from the database, and efficiently extracts relevant data using LIKE clauses and full-text search indexes.

[0597] Step 4:

[0598] The emotion engine analyzes the user's emotional data and prioritizes models that elicit positive emotional reactions. Emotional data is obtained by analyzing information collected from previous sessions, etc.

[0599] Step 5:

[0600] The server returns a list of models to the user in JSON format, which is then displayed on the user's device.

[0601] Step 6:

[0602] The user selects the model of interest from the returned list of models.

[0603] Step 7:

[0604] The device generates a request to send the selected model's ID to the server, which is also formatted as JSON.

[0605] Step 8:

[0606] The server retrieves detailed information about the model from the database. When retrieving, it uses a SELECT statement to extract detailed information about the model with the specified ID.

[0607] Step 9:

[0608] The server sends the detailed information it has obtained to the terminal in JSON format.

[0609] Step 10:

[0610] The device displays detailed information to the user, including a model description, performance metrics, price, and ratings.

[0611] Purchasing and using models

[0612] Step 1:

[0613] The user clicks the purchase button and enters payment information (such as credit card information). The entered payment information is encrypted to ensure security.

[0614] Step 2:

[0615] The device sends the purchase request and payment information to the server. The communication is encrypted using TLS / SSL during transmission.

[0616] Step 3:

[0617] The server then sends the payment information to an external payment processor via a secure channel using API integration.

[0618] Step 4:

[0619] The server receives the result of the payment process, and if successful, saves the success status along with the transaction ID.

[0620] Step 5:

[0621] The server issues a message confirming a successful purchase to the user along with access to the generative AI model, which is granted in the form of an additional session token.

[0622] Step 6:

[0623] Users access the download link or API endpoint to start using the purchased model.

[0624] Step 7:

[0625] The device sends a download request or API request to the server.

[0626] Step 8:

[0627] The server validates the user's session token to ensure access is permitted, and only if the validation is successful does it send the requested model file to the device.

[0628] Step 9:

[0629] The device unzips the received model file and begins using it in a local or cloud environment.

[0630] Model tuning and customization

[0631] Step 1:

[0632] If a user wishes to customize a purchased model, they open a customization request form within the system.

[0633] Step 2:

[0634] The terminal sends the customization request content (request details, specifications, dataset, etc.) entered by the user to the server.

[0635] Step 3:

[0636] The server forwards the received customization request to a specialized data scientist.

[0637] Step 4:

[0638] A data scientist receives the request and performs any necessary tuning. If any additional questions or clarifications are required, that information is also sent to the user via the server.

[0639] Step 5:

[0640] The server stores the completed tuning results in a database and notifies the user.

[0641] Step 6:

[0642] Users will be notified and will get new access information (download link, API endpoint, etc.) for the improved model.

[0643] Step 7:

[0644] The terminal sends an access request for the improved model to the server.

[0645] Step 8:

[0646] The server validates the user's session token to ensure access is permitted, and only if the validation is successful does it send the improved model file to the device.

[0647] Step 9:

[0648] The user unpacks the improved model they receive and starts using it in the required environment.

[0649] Monitoring and Maintenance

[0650] Step 1:

[0651] The server monitors the usage of the generated AI model in real time, including the frequency of API calls, error logs, and user operation history.

[0652] Step 2:

[0653] The server periodically collects and analyzes usage data, and if an anomaly is detected, an alert is generated and notifies the user.

[0654] Step 3:

[0655] Users receive notifications and can take action as needed, such as reporting a problem or suspending the model.

[0656] Step 4:

[0657] The server will check for any issues and take maintenance action as needed, including bug fixes and model retraining.

[0658] Step 5:

[0659] Once the server has completed its maintenance work, it will notify users that it is available again.

[0660] Incorporating an emotion engine

[0661] Step 1:

[0662] The server uses an emotion engine to collect emotional data in real time from the user's voice, text, facial expressions, etc. The collected data is temporarily stored in storage.

[0663] Step 2:

[0664] The server analyzes the collected emotion data and uses machine learning algorithms to determine the emotional state (e.g., stress, joy, sadness). This analysis can be done in batch or real-time.

[0665] Step 3:

[0666] Based on the emotion data acquired by the server, the system presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, models that are effective in reducing stress are prioritized.

[0667] Step 4:

[0668] The server's emotion engine periodically analyzes the emotion data and proposes or customizes models as needed. The analysis results are reflected in the next model proposal.

[0669] Step 5:

[0670] The user reviews the emotion-based suggested model and purchases or uses it. The suggested model is optimized for the user's current emotional state.

[0671] The above are the specific processing steps for implementing the present invention. The system provides domain-specific generative AI models, and can further improve the user experience by suggesting and customizing models based on the user's emotional state.

[0672] (Application example 2)

[0673] 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."

[0674] In the conventional AI model market, models based on user emotions were not provided, making it difficult to improve the user experience. Furthermore, it was difficult for users to find the optimal model under stressful circumstances, making it difficult to use the model efficiently. This resulted in low satisfaction with the purchase and use of AI models, limiting the overall effectiveness of the system.

[0675] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for building a marketplace that provides generative AI models specialized in specific fields, means for storing the generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and settling on the generative AI models, means for tuning the generative AI models in response to customization requests, means for monitoring the usage status of the generative AI models and performing necessary maintenance, means for recognizing the end user's emotions and proposing generative AI models based on the emotions, and means for analyzing emotion data and controlling the priority display of generative AI models. This makes it possible to provide personalized models according to the user's emotions, improving the user experience and dramatically improving the efficiency from model purchase to usage.

[0676] A "generative AI model" is an artificial intelligence model that is generated specifically for a specific field and exhibits high performance in specific tasks and data processing.

[0677] "Market" means an online marketplace provided for end users to search for, purchase, and use domain-specific generative AI models.

[0678] A "database" is a system for managing and storing data such as generative AI models, user information, transaction history, and emotional data.

[0679] "End User" means the ultimate user who accesses the system to utilize the Generative AI Model.

[0680] The "emotion engine" is a component that collects and analyzes emotional data in real time from end users' voices, text, facial expressions, etc.

[0681] A "Customization Request" is a request made by an End User to modify or adjust a Generative AI Model based on their specific needs or requirements.

[0682] "Tuning" is the process of adjusting and optimizing the performance and functionality of a generative AI model based on customization requirements.

[0683] "Priority display" is an operation that prominently presents generative AI models that are more appropriate and relevant to the user based on the results of emotional data analysis.

[0684] "Monitoring" is the process of monitoring the usage of generative AI models in real time and immediately detecting anomalies or problems.

[0685] "Maintenance" refers to the periodic management work of correcting and improving any problems or abnormalities that arise during use.

[0686] "Providing personalized models" means selecting and providing the optimal generative AI model based on the end user's individual emotional state and needs.

[0687] The present invention relates to a system for virtual stores that provides generative AI models specialized for specific fields. This system can recognize user emotions and propose optimal generative AI models based on those emotions. Specific embodiments for implementing this system are described below.

[0688] System configuration

[0689] 1. Building a Market:

[0690] The system will create a marketplace offering generative AI models in a virtual store, with a digital platform accessible to users that allows them to search, view details, purchase, and customize generative AI models.

[0691] 2. Use of databases:

[0692] The system uses a database to store and manage generative AI models, user information, transaction history, emotional data, etc. The database uses a common database management system such as MongoDB.

[0693] 3. User Device:

[0694] Users access the market using devices such as smartphones, tablets, and PCs, connecting to the system via a web browser or dedicated application.

[0695] 4. Server:

[0696] The server is the central component that receives user requests and performs the necessary processing. The server is built using a web framework such as Flask.

[0697] Specific processing steps

[0698] 1. User Registration and Authentication:

[0699] The server receives the information the user needs to create an account, validates it, and stores it in the database. When a user logs in, it checks the authentication information provided.

[0700] 2. Model exploration and selection:

[0701] The server receives a search query from the user's device and retrieves a list of relevant generative AI models from the database. The emotion engine analyzes the user's emotions and prioritizes generative AI models that evoke positive emotions.

[0702] 3. Use the Emotion Engine:

[0703] The server collects and analyzes emotional data from the user's touch points (e.g., microphone, camera) in real time. Based on the analysis results, it lists the generative AI model that is best suited to the user.

[0704] 4. Purchasing and using the model:

[0705] When a user purchases a generative AI model, the server processes the provided payment information. After purchase, the user can access the generative AI model and use it for download or via API.

[0706] 5. Processing customization requests:

[0707] The server forwards the customization request from the user to an expert (data scientist), who then provides a tuned generative AI model.

[0708] 6. Monitoring and Maintenance:

[0709] The server monitors the usage status of the generated AI model in real time and notifies the user if an abnormality is detected. Regular maintenance work is performed to maintain the suitability of the model for use.

[0710] Specific examples

[0711] 1. Example prompt:

[0712] "What field does the AI ​​model you're interested in relate to? Please be specific:"

[0713] "Are you feeling stressed? If so, would you suggest an AI model to help you relax?"

[0714] 2. Example of using the Emotion Engine:

[0715] When a user visits a virtual store, the emotion engine collects and analyzes the user's emotional data using a webcam and microphone. Based on the analysis results, the display priority of the model is controlled.

[0716] This invention makes it possible to provide an optimal generative AI model that corresponds to the user's emotions, thereby improving the user experience and enabling efficient model utilization.

[0717] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0718] Step 1: User Registration

[0719] Subject: User, Device, Server

[0720] Input: The user enters required information such as email address, password, and username on the account creation page.

[0721] Specific operation: The terminal sends the information entered by the user to the server as an HTTP request. The server validates the received information and, if there are no problems, saves the new user information in the database.

[0722] Output: The server returns a successful registration message to the user.

[0723] Step 2: User authentication

[0724] Subject: User, Device, Server

[0725] Input: The user enters their email address and password on the login page.

[0726] Specific operation: The terminal sends the entered authentication information to the server. The server searches the database for user information and performs authentication if there is a match.

[0727] Output: The server grants the user access to the market with a message of successful authentication.

[0728] Step 3: Model exploration

[0729] Subject: User, Device, Server

[0730] Input: A user enters a search query to explore generative AI models in a specific area within the market.

[0731] How it works: The device sends a search query to the server, which retrieves a list of relevant generative AI models from the database and analyzes the user's emotions using the emotion engine.

[0732] Output: The server returns a list of generated AI models that correspond to the user. Based on the emotion engine, models that match the user's emotions are displayed preferentially.

[0733] Step 4: Model selection

[0734] Subject: User, Device, Server

[0735] Input: The user selects the generative AI model of interest.

[0736] Specific operation: The device requests detailed information about the selected model. The server retrieves the details from the database and sends them to the device.

[0737] Output: The device displays detailed information about the generated AI model to the user.

[0738] Step 5: Purchase a model

[0739] Subject: User, Device, Server

[0740] Input: The user clicks the purchase button and enters payment information.

[0741] Specific operation: The device sends a purchase request and payment information to the server. The server processes the payment and, if successful, grants the user access to the generated AI model.

[0742] Output: The server sends the user a message confirming the purchase and providing an API key and download link.

[0743] Step 6: Use the model

[0744] Subject: User, Device, Server

[0745] Input: Uses a generative AI model purchased by the user.

[0746] Specific operation: The user accesses the generated AI model through their device. The server processes the API request and provides the required data and model to the user.

[0747] Output: The user uses the generative AI model for their own purposes.

[0748] Step 7: Customization Request

[0749] Subject: User, Device, Server

[0750] Input: The user details the customization requirements for the generative AI model.

[0751] How it works: The device sends a customization request to the server, which then forwards the request to a specialized data scientist.

[0752] Output: The server notifies the user when a new tuned model is available.

[0753] Step 8: Monitoring and Maintenance

[0754] Subject: Server

[0755] Input: Usage data for generative AI models

[0756] Specific operation: The server monitors the usage of the generative AI model in real time, analyzes the collected data, and if an abnormality is detected, notifies the user and performs necessary maintenance.

[0757] Output: The server evaluates the suitability of the model for use based on the results of the analysis of the usage data and makes corrections if necessary.

[0758] 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.

[0759] 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.

[0760] 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.

[0761] [Second embodiment]

[0762] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0763] 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.

[0764] 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).

[0765] 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.

[0766] 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.

[0767] 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).

[0768] 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.

[0769] 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.

[0770] 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.

[0771] 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.

[0772] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0773] 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."

[0774] A system for providing generative AI models specialized for specific fields is configured as follows.

[0775] System Overview

[0776] This system builds a marketplace that provides generative AI models specialized in specific fields, and allows users to access the marketplace to search, purchase, and use models. The system mainly consists of the following components:

[0777] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[0778] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes databases, authentication systems, and payment systems.

[0779] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, and transaction history.

[0780] Explanation of program processing

[0781] User Registration and Authentication

[0782] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[0783] Terminal: Sends the entered information to the server as an HTTP request.

[0784] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[0785] Server: Returns a successful registration message to the user.

[0786] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[0787] Model Exploration and Selection

[0788] Users: Explore generative AI models optimized for specific domains (e.g., healthcare, finance, etc.).

[0789] Device: Sends the user's search query to the server.

[0790] Server: Retrieves a list of relevant generative AI models from the database and returns them to the user.

[0791] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[0792] Server: Retrieves detailed information from the database and sends it to the device.

[0793] Terminal: Display detailed information to the user.

[0794] Purchasing and using models

[0795] User: Select the model they wish to purchase and enter their payment information to complete the purchase.

[0796] Terminal: Sends the entered payment information to the server.

[0797] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[0798] Users: Download purchased models or access them through API.

[0799] Model tuning and customization

[0800] User: If a purchased model needs customization, the user sends a request to the server with their specific requirements.

[0801] Server: Receives requests and relays them to expert data scientists.

[0802] Server: Data scientists tune the model and generate a new model.

[0803] Server: When a new model is ready, it notifies the user and provides access.

[0804] Monitoring and Maintenance

[0805] Server: Monitors the usage of the generative AI model in real time.

[0806] Server: Collects and analyzes usage data and notifies users if anomalies are detected, such as when a model responds excessively slowly or when the error rate is high.

[0807] Server: Once an abnormality is confirmed, maintenance work will be carried out and users will be notified that the service is available again once the problem has been fixed.

[0808] Specific examples

[0809] 1. User Registration and Authentication Example

[0810] Researchers at medical institutions register an account to use the system for the first time.

[0811] Researchers log in and search for generative AI models to find AI models specialized for pathological diagnosis.

[0812] 2. Example of model purchase

[0813] Researchers check detailed information about the pathology diagnostic AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[0814] After purchase, researchers can download the model and begin using it in their actual research.

[0815] 3. Model Tuning Example

[0816] A researcher requests customization, requesting detailed tuning based on clinical data.

[0817] Specialized data scientists will handle the process and provide newly tuned models.

[0818] This invention makes it possible to safely and effectively provide generative AI models specialized for specific fields, enabling users to expect high effectiveness in those fields. In addition, the ease of customization and maintenance improves user convenience.

[0819] The processing flow will be explained below.

[0820] User Registration and Authentication

[0821] Step 1:

[0822] User: Visit the Market and open the account creation page.

[0823] Step 2:

[0824] Device: Enter the required information (email address, password, username, etc.).

[0825] Step 3:

[0826] Terminal: Sends the entered information to the server as an HTTP request.

[0827] Step 4:

[0828] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[0829] Step 5:

[0830] Server: Returns a successful registration message to the user.

[0831] Step 6:

[0832] User: Opens the login page and enters their email address and password.

[0833] Step 7:

[0834] Terminal: Sends the entered authentication information to the server.

[0835] Step 8:

[0836] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[0837] Step 9:

[0838] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[0839] Model Exploration and Selection

[0840] Step 1:

[0841] User: Select a specific sector within the market (e.g., healthcare, finance, etc.).

[0842] Step 2:

[0843] On your device: Enter your search query in the search bar within Market.

[0844] Step 3:

[0845] Device: Sends a search query to the server.

[0846] Step 4:

[0847] Server: Obtain a list of relevant generative AI models from the database.

[0848] Step 5:

[0849] Server: Returns the retrieved model list to the user.

[0850] Step 6:

[0851] User: Select the model of interest from the returned list of models.

[0852] Step 7:

[0853] User: Sends a request to view detailed information about a selected model.

[0854] Step 8:

[0855] Server: Retrieves detailed information from the database and sends it to the device.

[0856] Step 9:

[0857] Terminal: Display detailed information to the user.

[0858] Purchasing and using models

[0859] Step 1:

[0860] User: Clicks the purchase button and enters payment information.

[0861] Step 2:

[0862] Terminal: Sends purchase request and payment information to the server.

[0863] Step 3:

[0864] Server: Processes the payment, and if successful, sends the user a purchase confirmation message and grants them access to the model.

[0865] Step 4:

[0866] Users: Download purchased models or access them via API.

[0867] Model tuning and customization

[0868] Step 1:

[0869] Users: If you require customization of the purchased model, please describe your request in detail.

[0870] Step 2:

[0871] Device: Sends a customization request to the server.

[0872] Step 3:

[0873] Server: Forwards the request to an expert data scientist.

[0874] Step 4:

[0875] Server: Data scientists perform tuning and create new generative AI models.

[0876] Step 5:

[0877] Server: Saves the new generative AI model in the database and notifies the user.

[0878] Step 6:

[0879] User: Use the improved model with the new access information.

[0880] Monitoring and Maintenance

[0881] Step 1:

[0882] Server: Monitors the usage of the generative AI model in real time.

[0883] Step 2:

[0884] Server: Collects and analyzes usage data.

[0885] Step 3:

[0886] Server: If an anomaly is detected, a notification is sent to the user.

[0887] Step 4:

[0888] Users: Receive notifications and take action as needed.

[0889] Step 5:

[0890] Server: If a problem is identified, maintenance work is carried out to fix it.

[0891] Step 6:

[0892] Server: Once the fix is ​​complete, notify users that it is available again.

[0893] In this way, the system provides generative AI models specialized for specific fields and can be effectively operated to meet the diverse needs of users.

[0894] Example 1

[0895] 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."

[0896] Platforms that provide generative AI models specialized in specific fields are required to build an environment where users can efficiently search, purchase, and use generative AI models. However, many current systems lack sufficient functionality for user authentication, search query analysis, customization request response, usage monitoring, and anomaly detection, making it difficult to improve user convenience and reliability.

[0897] 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.

[0898] In this invention, the server includes means for building a platform that provides generative AI models specialized in specific fields, means for storing generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for generative AI models, means for adjusting generative AI models in response to customization requests, means for monitoring the usage status of generative AI models and performing necessary maintenance work, means for users to log in and input authentication information, and means for analyzing user search queries and suggesting appropriate generative AI models. This enables users to efficiently search for, purchase, customize, and use generative AI models.

[0899] "Specific fields" refer to areas with specific expertise or needs, such as medicine, finance, or education.

[0900] "Generative AI models" refer to artificial intelligence models that generate new data and content using techniques such as generative adversarial networks (GANs) and natural language generation (NLG).

[0901] "Platform" refers to the infrastructure that provides an online environment for users to search for, purchase, and use generative AI models.

[0902] "Database" refers to a system for systematically storing and managing generative AI models, related user information, transaction history, etc.

[0903] "End User" refers to the ultimate user who utilizes the generative AI model to provide a specific application or service.

[0904] A "search query" refers to a keyword or phrase entered by a user when searching for a generative AI model.

[0905] "Customization Request" means a request submitted by a User to adjust or refine a Generative AI Model based on their specific needs.

[0906] "Maintenance work" refers to regular checks and maintenance work to ensure that generative AI models function properly.

[0907] "Login" refers to the authentication procedure required for a user to access a system.

[0908] "Authentication Information" refers to information such as email address and password that a User provides to prove access to a System.

[0909] "Search query analysis" refers to the process of analyzing a search query entered by a user and proposing the optimal generative AI model.

[0910] The system of the present invention builds a platform that provides generative AI models specialized in specific fields, and allows end users to search, purchase, customize, and use generative AI models using this platform. The main components of the system are as follows:

[0911] 1. User Device

[0912] A device that users access and operate. It can be a PC, smartphone, tablet, etc. Users access the system through their device to register an account, log in, search for models, view detailed information, make purchases, and request customization.

[0913] 2. Server

[0914] The server is the central component that manages and operates the entire system. It stores generative AI models in a database and includes an authentication system for user authentication, a search query analysis system, a payment system, and a system for responding to customization requests. It also provides API endpoints for users to access. The server communicates with the database and returns appropriate information based on the user's request.

[0915] 3. Database

[0916] This is a system for storing data such as generative AI models, user information, and transaction history. This database uses database systems such as PostgreSQL and Elasticsearch, particularly to speed up searches and maintain data consistency.

[0917] For example, a user accesses a market website and enters the required information (email address, password, username, etc.) on the account registration page. The device sends this information to the server, which validates it and stores it in the database. When the user enters their authentication information on the login page, the server authenticates the user and, if successful, allows them to access the market.

[0918] Next, the user searches for generative AI models suitable for a specific field (e.g., medicine or finance). The device sends the search query to the server, which retrieves and returns a list of relevant models from the database. The user selects the model of interest and requests that detailed information be displayed. The server retrieves the detailed information from the database and sends it to the device, allowing the user to purchase or customize the model.

[0919] Furthermore, if a user requests customization, the request is forwarded to the server and handled by a specialized data scientist. Once customization is complete, a new model is generated and provided to the user. The server also monitors the usage of the generated AI model in real time and takes appropriate action if an abnormality is detected.

[0920] Prompt Sentence Examples

[0921] 1. Researchers at medical institutions must register an account to use the system for the first time.

[0922] 2. Purchase and download the AI ​​model to be used for pathology diagnosis.

[0923] 3. Customize the pathology diagnostic AI model you purchased based on clinical data.

[0924] In this way, the system effectively provides domain-specific generative AI models, enhancing user convenience.

[0925] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0926] Step 1: User Registration

[0927] Users: Visit the Market registration page and enter your name, email address, and password.

[0928] Input: Name, email address, and password entered by the user into the form.

[0929] Output: Registration successful message.

[0930] Terminal: Sends user input information to the server in JSON format.

[0931] Input: Information entered by the user.

[0932] Output: HTTP POST request in JSON format.

[0933] Server: Validate the received information using a validation library (e.g. Joi) and, if there are no problems, save the user information to the PostgreSQL database.

[0934] Input: User information in JSON format.

[0935] Output: The new user information is saved in the database.

[0936] Server: Returns a successful registration message to the user in JSON format.

[0937] Input: User information saved successfully.

[0938] Output: Registration successful message.

[0939] Step 2: User Login

[0940] User: Access the login page and enter the registered email address and password.

[0941] Enter your email address and password.

[0942] Output: Login request.

[0943] Terminal: Sends the user's authentication information to the server in a JSON-formatted HTTP POST request.

[0944] Input: The credentials entered by the user.

[0945] Output: HTTP POST request in JSON format.

[0946] Server: Validates the received authentication information using a password authentication system (e.g. bcrypt) and issues a JWT token if it matches.

[0947] Input: User credentials.

[0948] Output: JWT token.

[0949] Server: Sends the JWT token to the user terminal.

[0950] Input: JWT token.

[0951] Output: User authentication success message and JWT token.

[0952] Step 3: Model Search

[0953] Users: Find generative AI models optimized for specific domains through search queries.

[0954] Input: Search query.

[0955] Output: The search request.

[0956] Terminal: Sends the user's query to the server as an HTTP GET request in JSON format.

[0957] Input: Search query.

[0958] Output: HTTP GET request in JSON format.

[0959] Server: Uses a search engine such as Elasticsearch to retrieve the relevant generative AI model from the database.

[0960] Input: Search query.

[0961] Output: A list of generative AI models.

[0962] Server: Returns search results to the user's device.

[0963] Input: A list of generative AI models.

[0964] Output: Search results.

[0965] Step 4: View model details

[0966] User: Select the model they are interested in and request more information.

[0967] Input: Model ID.

[0968] Output: Model details request.

[0969] Device: Send an HTTP GET request in JSON format containing the model ID to the server.

[0970] Input: Model ID.

[0971] Output: HTTP GET request in JSON format.

[0972] Server: Obtains detailed information about the selected generative AI model from the database and sends it to the user's device.

[0973] Input: Model ID.

[0974] Output: Detailed information about the model.

[0975] Terminal: Displays the received detailed information to the user.

[0976] Input: Model details.

[0977] Output: The detailed information displayed to the user.

[0978] Step 5: Purchase a model

[0979] User: Selects the model they want to purchase and enters their payment information to begin the checkout process.

[0980] Input: Payment information.

[0981] Output: Purchase request.

[0982] Terminal: Sends the user's payment information to the server via an HTTPS POST request in JSON format.

[0983] Input: Payment information.

[0984] Output: HTTPS POST request in JSON format.

[0985] Server: Process the payment using the Stripe API, and if successful, record the purchase confirmation message and model access rights in the database.

[0986] Input: Payment information.

[0987] Output: Purchase successful message and updated access rights.

[0988] Server: Sends a purchase confirmation message and access right information to the user terminal.

[0989] Input: Purchase success message and access rights information.

[0990] Output: Notification to user terminal.

[0991] Step 6: Use the model

[0992] Users: Access or download purchased models via API.

[0993] Input: Access request.

[0994] Output: Model use.

[0995] Server: Receives requests to access a model, performs appropriate authentication, and returns the model file or issues an API key.

[0996] Input: Access request.

[0997] Output: Model file or API key.

[0998] Terminal: Provide the received model file or API key to the user.

[0999] Input: Model file or API key.

[1000] Output: The model is made available to the user.

[1001] (Application example 1)

[1002] 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."

[1003] Modern factories require efficient and rapid optimization of manufacturing processes. It is difficult to easily obtain and use generative AI models suited to specific manufacturing environments. Furthermore, customization of generative AI models and effective monitoring after their deployment are required. To address these challenges, an effective means is needed to easily provide AI models for optimizing the operation of factory robots, and to customize and maintain those models.

[1004] 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.

[1005] In this invention, the server includes means for building a marketplace that provides generative AI models specialized in specific fields, means for storing the generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for the generative AI models, means for tuning the generative AI models in response to customization requests, means for monitoring the usage status of the generative AI models and performing necessary maintenance, means for end users to select, download, or integrate generative AI models to be applied to factory robots, and means for optimizing the operation of the factory robots for specific manufacturing processes. This enables end users to easily obtain generative AI models suitable for their manufacturing environment and effectively apply and operate them.

[1006] A "generative AI model" is an algorithmic model of artificial intelligence that is generated for a specific field or application, and exhibits performance optimized for that field.

[1007] "Market" means the online platform where end users can find, purchase, customize, download, and integrate Generative AI Models.

[1008] A "database" is an information recording system for managing and storing generative AI models, user information, transaction history, etc.

[1009] "End users" are the final users of generative AI models, such as factory workers and managers.

[1010] "Customization Request" means a request by an End User to tailor a Generative AI Model based on their specific needs and requirements.

[1011] "Tuning" refers to the process of adjusting a generative AI model to optimize it for the end user's specific use and environment.

[1012] "Monitoring" refers to the act of monitoring the usage and performance of generative AI models in real time.

[1013] "Maintenance" refers to maintenance work to maintain the performance and reliability of the generative AI model, including fixing bugs and making improvements.

[1014] A "factory robot" is a mechanical device used to automate specific tasks in manufacturing sites, and its operations are controlled by a program.

[1015] "Manufacturing process" refers to the series of operations or steps that transform raw materials into a finished product, including welding, assembly, quality inspection, etc.

[1016] "Integration" refers to the act of end users applying purchased generative AI models to their own systems or machinery (e.g., factory robots) to make them work together.

[1017] This invention relates to a system that enables end users to easily search, purchase, customize, download, and integrate generative AI models suited to specific manufacturing processes. This system is primarily composed of the following hardware and software:

[1018] Hardware and Software Use Cases

[1019] Hardware: A factory robot (e.g., a generic robotic arm) and the smartphone or tablet (iOS or Android device) that controls it.

[1020] software:

[1021] Robot control software (e.g., Robot Operating System (ROS)).

[1022] A cloud service (e.g., AWS or Google Cloud) as a model management server.

[1023] The application for operation on smart devices will be developed using React Native.

[1024] Program processing explanation

[1025] In this invention, the server, terminal, and user work together to search, select, purchase, customize, and use generative AI models. The specific processing steps are as follows:

[1026] 1. User Registration and Authentication

[1027] Users open the app on their smartphone and enter the required information on the account creation page.

[1028] The terminal transmits the input information to the server as an HTTP request.

[1029] The server validates the received information and, if there are no problems, saves the new user information in the database and returns a message to the user confirming successful registration.

[1030] 2. Model Search and Selection

[1031] Users enter a search query to explore generative AI models specific to their manufacturing process.

[1032] The terminal transmits the user's search query to the server.

[1033] The server retrieves a list of relevant generative AI models from the database and returns it to the user.

[1034] The user selects a model of interest from the returned list and requests that its detailed information be displayed.

[1035] The server retrieves the detailed information from the database and sends it to the terminal.

[1036] The terminal displays detailed information to the user.

[1037] 3. Purchasing and Using the Model

[1038] Users select the generative AI model they want to purchase and enter their payment information.

[1039] The terminal transmits the entered payment information to the server.

[1040] The server processes the payment and, if successful, gives the user a purchase confirmation message and access to the model.

[1041] Users can download the purchased model or integrate it into their robot control system via API.

[1042] 4. Tuning and customizing the model

[1043] When a user desires tuning specialized for a specific production line, the user writes a customization request and sends it to the server.

[1044] The server receives the request and relays it to the data scientist.

[1045] The server will notify the user and grant access as soon as a new tuned model is available.

[1046] 5. Monitoring and Maintenance

[1047] The server monitors the usage of the generative AI model in real time.

[1048] The server collects and analyzes usage data and notifies the user if an anomaly is detected, such as a robot slowing down or an increased error rate.

[1049] The server will perform maintenance when an abnormality is detected and notify users that the service is available again once the problem has been fixed.

[1050] Specific examples

[1051] For example, if an end user wants to apply an AI model to a welding process in a factory, they might enter a prompt like this:

[1052] "Optimize the welding seam angle and speed used by this robotic arm."

[1053] This allows the generative AI model to optimize the parameters of the welding process, reducing error rates and improving production speed.

[1054] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1055] Step 1: User Registration and Authentication

[1056] User: Opens the smartphone app and enters the required information (email address, password, factory ID, etc.) on the account creation page.

[1057] Terminal: Sends the entered information to the server as an HTTP request.

[1058] Server: Validates the received information and, if there are no problems, saves the new user information to the database. If the save is successful, generates a message indicating successful registration and returns it to the terminal. Examples of validation include detecting invalid email addresses and passwords that are too short.

[1059] Step 2: Log in

[1060] User: Enters email address and password on the login page and clicks the login button.

[1061] Terminal: Sends the entered information to the server as an HTTP request.

[1062] Server: Searches for the corresponding user information in the database and compares it with the authentication information. If it matches, it generates a login success message and session information and sends them to the terminal.

[1063] Step 3: Model exploration

[1064] User: Enter a specific manufacturing process (e.g., welding, assembly) into the app's search bar and press the search button.

[1065] Terminal: Sends the search query as an HTTP request to the server.

[1066] Server: Searches the database for a list of relevant generative AI models, generates a list of relevant models, and sends it to the device. Specifically, it filters the models by tags and keywords related to the manufacturing process.

[1067] Step 4: Model selection and detailed display

[1068] User: Selects the generative AI model of interest from the returned list and requests more information about it.

[1069] Device: Send the model ID to the server as an HTTP request.

[1070] Server: Retrieves detailed information about the selected model from the database, generates detailed information, and sends it to the terminal. The detailed information includes the model's scope of application, expected results, and usage methods.

[1071] Step 5: Purchase and pay for the model

[1072] User: Select the generative AI model they want to purchase and enter their payment information (credit card number, security code, etc.).

[1073] Terminal: Sends payment information to the server as an HTTP request.

[1074] Server: Works with the payment processing system to perform payment, and if successful, generates a purchase confirmation message and access rights for the model and sends them to the terminal. If unsuccessful, generates an error message and sends it to the terminal.

[1075] Step 6: Download and integrate the model

[1076] User: Presses a button to download the purchased generative AI model.

[1077] Device: Sends a model download request to the server.

[1078] Server: Retrieves the corresponding model file from the database, generates a download link, and sends it to the device.

[1079] Users: Click the download link and integrate the model file into their robot control system. Specifically, they load the model using software such as ROS and set parameters to optimize the robot's behavior.

[1080] Step 7: Tune and customize the model

[1081] User: If a user wants specialized tuning for a specific production line, he or she enters a customization request and sends it to the server.

[1082] Terminal: Sends the customization request to the server as an HTTP request.

[1083] Server: Receives requests and forwards them to data scientists, who tune the model based on the requests and generate new versions of the model.

[1084] Server: Saves the new tuned model to the database and notifies the user of the update.

[1085] Step 8: Monitoring and Maintenance

[1086] Server: Collects logging and performance metrics to monitor the usage of generative AI models in real time.

[1087] Server: If abnormal behavior (e.g., delayed response, increased error rate) is detected, the server identifies the anomaly based on an anomaly detection algorithm and notifies the user. Statistical anomaly detection techniques and machine learning algorithms are used for anomaly detection.

[1088] Server: If an anomaly is detected, the maintenance team is notified and the model is available again once the fix is ​​complete. Continuous monitoring ensures the reliability and effectiveness of the model.

[1089] 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.

[1090] The system for providing generative AI models specialized for specific fields is configured as follows: This system also incorporates an emotion engine that recognizes user emotions.

[1091] System Overview

[1092] This system creates a marketplace that provides generative AI models specialized in specific fields, allowing users to access the marketplace and search, purchase, and use models. It also has the ability to recognize user emotions and suggest models and interact with them based on those emotions.

[1093] The system mainly consists of the following components:

[1094] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[1095] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes a database, authentication system, payment system, emotion engine, etc.

[1096] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, transaction history, and emotional data.

[1097] 4. Emotion Engine: A component for recognizing and analyzing user emotions.

[1098] Explanation of program processing

[1099] User Registration and Authentication

[1100] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[1101] Terminal: Sends the entered information to the server as an HTTP request.

[1102] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[1103] Server: Returns a successful registration message to the user.

[1104] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[1105] Model Exploration and Selection

[1106] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[1107] Device: Sends the user's search query to the server.

[1108] Server: Retrieves a list of applicable generative AI models from the database and returns it to the user. The emotion engine recognizes the user's emotions and prioritizes models that evoke positive emotional responses.

[1109] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[1110] Server: Retrieves detailed information from the database and sends it to the device.

[1111] Terminal: Display detailed information to the user.

[1112] Purchasing and using models

[1113] User: Clicks the purchase button and enters payment information.

[1114] Terminal: Sends purchase request and payment information to the server.

[1115] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[1116] Users: Download purchased models or access them via API.

[1117] Model tuning and customization

[1118] Users: If you require customization of the purchased model, please describe your request in detail.

[1119] Device: Sends a customization request to the server.

[1120] Server: Forwards the request to an expert data scientist.

[1121] Server: Data scientists perform tuning and create new generative AI models.

[1122] Server: Saves the new generative AI model in the database and notifies the user.

[1123] User: Use the improved model with the new access information.

[1124] Monitoring and Maintenance

[1125] Server: Monitors the usage of the generative AI model in real time.

[1126] Server: Collects and analyzes usage data.

[1127] Server: If an anomaly is detected, a notification is sent to the user.

[1128] Users: Receive notifications and take action as needed.

[1129] Server: If a problem is identified, maintenance work is carried out to fix it.

[1130] Server: Once the fix is ​​complete, notify users that it is available again.

[1131] Incorporating an emotion engine

[1132] Server: Builds an emotion engine and collects emotion data in real time from users' voices, text, facial expressions, etc.

[1133] Server: The emotion engine analyzes the collected data and determines the emotional state.

[1134] Server: Based on the acquired emotion data, the server presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, the server prioritizes models that are effective in reducing stress.

[1135] Server: The emotion engine periodically analyzes the emotion data and suggests or customizes models as needed.

[1136] User: Review the sentiment-based recommendation model and make a purchase or use it.

[1137] Specific examples

[1138] 1. User Registration and Authentication Example

[1139] A pharmaceutical researcher from a healthcare company registers an account to use the system for the first time.

[1140] Researchers log in and search generative AI models to find AI models specialized for specific diseases.

[1141] 2. Example of model purchase

[1142] Researchers check detailed information about the disease-specific AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[1143] After purchase, researchers can download the model and begin using it in their research.

[1144] 3. Model Tuning Example

[1145] A researcher requests customization, requesting detailed tuning based on clinical data.

[1146] Specialized data scientists will handle the process and provide newly tuned models.

[1147] 4. Example of an Emotion Engine

[1148] As researchers use the system, the emotion engine analyzes their context and recognizes when they are feeling stressed.

[1149] Based on the emotion engine, we present effective guidelines for stress reduction and propose a model to improve researchers' work efficiency.

[1150] This invention enables the safe and effective provision of generative AI models specialized for specific fields. In addition, by incorporating an emotion engine, it is possible to provide more personalized services according to the user's emotional state.

[1151] The processing flow will be explained below.

[1152] User Registration and Authentication

[1153] Step 1:

[1154] User: Visit the Market and open the account creation page.

[1155] Step 2:

[1156] Device: Enter the required information (email address, password, username, etc.).

[1157] Step 3:

[1158] Terminal: Sends the entered information to the server as an HTTP request.

[1159] Step 4:

[1160] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[1161] Step 5:

[1162] Server: Returns a successful registration message to the user.

[1163] Step 6:

[1164] User: Opens the login page and enters their email address and password.

[1165] Step 7:

[1166] Terminal: Sends the entered authentication information to the server.

[1167] Step 8:

[1168] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[1169] Step 9:

[1170] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[1171] Model Exploration and Selection

[1172] Step 1:

[1173] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[1174] Step 2:

[1175] On your device: Enter your search query in the search bar within Market.

[1176] Step 3:

[1177] Device: Sends a search query to the server.

[1178] Step 4:

[1179] Server: Obtain a list of relevant generative AI models from the database.

[1180] Step 5:

[1181] Server: Returns the acquired model list to the user device.

[1182] Step 6:

[1183] User: Select the model of interest from the returned list of models.

[1184] Step 7:

[1185] User: Sends a request to view detailed information about a selected model.

[1186] Step 8:

[1187] Server: Retrieves detailed information from the database and sends it to the device.

[1188] Step 9:

[1189] Terminal: Display detailed information to the user.

[1190] Step 10:

[1191] Emotion engine: Analyzes the user's facial expressions, voice, text, etc. to obtain emotional data.

[1192] Step 11:

[1193] Server: Based on the emotion data, add the generative AI model that is suitable for the user to the suggestion list.

[1194] Purchasing and using models

[1195] Step 1:

[1196] User: Clicks the purchase button and enters payment information.

[1197] Step 2:

[1198] Terminal: Sends purchase request and payment information to the server.

[1199] Step 3:

[1200] Server: Processes the payment, and if successful, sends the user a purchase confirmation message and grants them access to the model.

[1201] Step 4:

[1202] Users: Download purchased models or access them through the API.

[1203] Model tuning and customization

[1204] Step 1:

[1205] Users: If you need customization for the model you purchased, please submit your specific request.

[1206] Step 2:

[1207] Terminal: Forwards the customization request to the server.

[1208] Step 3:

[1209] Server: Forwards the request to an expert data scientist.

[1210] Step 4:

[1211] Server: Data scientists tune the model and generate a new generative AI model.

[1212] Step 5:

[1213] Server: Saves the newly tuned generative AI model in the database and notifies the user.

[1214] Step 6:

[1215] User: Use the improved model with the new access information.

[1216] Monitoring and Maintenance

[1217] Step 1:

[1218] Server: Monitors the usage of the generative AI model in real time.

[1219] Step 2:

[1220] Server: Collects and analyzes usage data.

[1221] Step 3:

[1222] Server: If an anomaly is detected, a notification is sent to the user.

[1223] Step 4:

[1224] Users: Receive notifications and take action as needed.

[1225] Step 5:

[1226] Server: If a problem is identified, maintenance work is carried out to fix it.

[1227] Step 6:

[1228] Server: Once the fix is ​​complete, notify users that it is available again.

[1229] Incorporating an emotion engine

[1230] Step 1:

[1231] Server: Builds an emotion engine and collects emotion data in real time from users' voices, text, facial expressions, etc.

[1232] Step 2:

[1233] Server: The emotion engine analyzes the collected data and determines the emotional state.

[1234] Step 3:

[1235] Server: Based on the acquired emotion data, presents the user with a list of optimal generative AI models.

[1236] Step 4:

[1237] Server: Adjusts suggestions and interactions to the user based on emotion data. For example, if the user is feeling stressed, it makes suggestions to help them relax.

[1238] Specific examples

[1239] User registration and authentication examples

[1240] A researcher opens the account registration page.

[1241] The researcher enters the required information and completes the registration.

[1242] The researcher logs in using their email address and password.

[1243] Specific examples of model purchases

[1244] Researchers select a disease-specific AI model and view detailed information.

[1245] The researcher decides to purchase and enters payment information.

[1246] Researchers can download the model and use it in their research.

[1247] Example of model tuning

[1248] The researcher details and submits the customization request.

[1249] The server forwards the request to an expert data scientist.

[1250] Data scientists tune the model and provide a new model.

[1251] Examples of emotion engines

[1252] While researchers are using the system, an emotion engine detects their stress levels.

[1253] The server proposes a model suitable for stress reduction.

[1254] Researchers will follow the suggestions and use the model to improve operational efficiency.

[1255] This invention makes it possible to provide generative AI models specialized for specific fields while providing individual services according to the user's emotional state. The incorporation of an emotion engine is expected to improve the user experience and lead to more effective use.

[1256] Example 2

[1257] 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."

[1258] In recent years, the use of generative AI models has expanded in various fields, but there is a lack of systems that efficiently provide generative AI models that can meet specialized requirements.In addition, the lack of models that adapt to the user's emotional state often leads to a decrease in user satisfaction and hinders the effective use of the models.

[1259] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for building a market that provides generative AI models specialized in a specific field, a means for storing the generative AI models in a database, and a means for analyzing user emotion data and proposing an optimal generative AI model based on the analyzed data. This makes it possible to efficiently provide specialized generative AI models and propose and use models that are adapted to the user's emotional state.

[1260] "Market" is an online platform for providing domain-specific generative AI models.

[1261] A "generative AI model" is an artificial intelligence model that is generated based on data from a specific field and automates or assists with various tasks in that field.

[1262] A "database" is a system for storing generative AI models, user information, emotional data, etc.

[1263] "End User" means an individual or legal entity that uses the System to search for, purchase, and use Generative AI Models.

[1264] "Search Means" means means that provide the ability for an End User to search for a particular Generative AI Model.

[1265] "Means for displaying detailed information" refers to means for providing a function for end users to display detailed information (e.g., performance indicators, price, usage methods, etc.) about the generated AI model in which they are interested.

[1266] "Means for purchasing and payment" refers to means that provide the functionality for processing payments when end users purchase generative AI models.

[1267] A "customization request" is a request by an end user for additional functionality or tuning of a generative AI model that they have purchased.

[1268] "Means for tuning" means means for providing the ability to adjust or improve a generative AI model in response to end-user customization requests.

[1269] "Monitoring means" refers to a means for monitoring the usage of the generated AI model in real time and providing the functionality to detect anomalies and collect data.

[1270] "Maintenance measures" are measures for carrying out corrective work or maintenance to resolve abnormalities or problems detected through monitoring.

[1271] "Emotional data" is data that indicates the user's emotional state and is collected from voice, text, facial expressions, etc.

[1272] The "means for analyzing emotional data" is a means for analyzing collected emotional data and providing a function for determining the emotional state of the user.

[1273] The "emotion engine" is a system component that proposes and adjusts models based on the user's emotional state.

[1274] The present invention relates to a system that provides generative AI models specialized in specific fields. The system allows users to access, search, purchase, and use specialized generative AI models, and further has the function of analyzing user emotion data and proposing optimal generative AI models based on the data.

[1275] System Overview

[1276] The system consists of the following components:

[1277] 1. User terminal: A device that a user accesses and operates. Examples include PCs, smartphones, tablets, etc.

[1278] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. The server includes a database, authentication system, payment system, emotion engine, etc.

[1279] 3. Database: Stores data such as domain-specific generative AI models, user information, transaction history, and sentiment data.

[1280] 4. Emotion Engine: A component for recognizing and analyzing user emotions.

[1281] User Registration and Authentication

[1282] When a user accesses the market, they first enter the required information on the account creation page. The device sends this information to the server as an HTTP request. The server validates the received information and, if there are no problems, saves the new user information in the database. The user then enters their authentication information on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[1283] Model Exploration and Selection

[1284] When a user searches the market for a generative AI model optimized for a specific field, the device sends a search query to the server. The server retrieves a list of relevant generative AI models from the database and returns it to the user. The emotion engine recognizes the user's emotions and prioritizes models that evoke positive emotional reactions. The user selects a model of interest from the returned list of models and requests that its detailed information be displayed on the device. The server retrieves the detailed information from the database and sends it to the device. The device displays this information to the user.

[1285] Purchasing and using models

[1286] When a user purchases a model, they click the purchase button and enter their payment information. The device sends this request and payment information to the server. The server processes the payment through an external payment gateway, and if successful, issues the user a purchase confirmation message and access rights to the model. The user can then download the purchased model or access it via API.

[1287] Model tuning and customization

[1288] If a user wants to customize a model, they send a detailed request from their device to the server. The server then forwards the request to a specialized data scientist. Once the data scientist performs tuning and a new generative AI model is created, the server stores it in a database and notifies the user. The user can then use the improved model with the new access information.

[1289] Monitoring and Maintenance

[1290] The server monitors the usage of the generative AI model in real time, collecting and analyzing usage data. If an abnormality is detected, a notification is sent to the user. The user receives the notification and can take action as necessary. If a problem is confirmed, the server performs maintenance work, and once the problem is fixed, the server notifies the user that the model is available again.

[1291] Incorporating an emotion engine

[1292] The emotion engine allows the server to collect and analyze emotional data from the user's voice, text, facial expressions, etc. in real time. Based on the acquired emotional data, the server presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, models that are effective in reducing stress will be prioritized. The emotion engine regularly analyzes the emotional data and suggests or customizes models as needed.

[1293] Specific examples

[1294] 1. User Registration and Authentication Example

[1295] A pharmaceutical researcher from a healthcare company registers an account to use the system for the first time.

[1296] Researchers log in and search generative AI models to find AI models specialized for specific diseases.

[1297] 2. Example of model purchase

[1298] Researchers check detailed information about the disease-specific AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[1299] After purchase, researchers can download the model and begin using it in their research.

[1300] 3. Model Tuning Example

[1301] A researcher requests customization, requesting detailed tuning based on clinical data.

[1302] Specialized data scientists will handle the process and provide newly tuned models.

[1303] 4. Example of an Emotion Engine

[1304] As researchers use the system, the emotion engine analyzes their context and recognizes when they are feeling stressed.

[1305] Based on the emotion engine, we present effective guidelines for stress reduction and propose a model to improve researchers' work efficiency.

[1306] As a result, the present invention can safely and effectively provide generative AI models specialized for specific fields, while also providing personalized services that correspond to the user's emotional state.

[1307] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1308] User Registration and Authentication

[1309] Step 1:

[1310] The user accesses the market and enters the required information (email address, password, username, etc.) on the account creation page. The entered information is displayed on the user's device, and once confirmation is complete, the user presses the send button.

[1311] Step 2:

[1312] The device sends the entered information to the server as an HTTP request, which contains the entered information and is formatted in JSON.

[1313] Step 3:

[1314] The server validates the information it receives. For example, it checks the format of the email address or the strength of the password. This validation is performed using regular expressions or rule-based checks. If the input is found to be valid, it proceeds.

[1315] Step 4:

[1316] The server stores the successfully validated information in the database. The password is hashed and stored securely along with the user information. The storage process uses an SQL query.

[1317] Step 5:

[1318] The server returns a message to the user confirming successful registration. The message is structured in JSON format and sent to the terminal for display.

[1319] Step 6:

[1320] The user again enters their authentication information (email address, password) on the login page. The entered information is again sent by the terminal to the server as an HTTP request.

[1321] Step 7:

[1322] The server checks the received authentication information against its database, specifically to ensure that the hashed password matches. If the check is successful, it issues a session token to the user and returns a response containing that session token.

[1323] Step 8:

[1324] Users can access the market using a session token, which is stored in a browser cookie and used to simplify authentication in future visits.

[1325] Model Exploration and Selection

[1326] Step 1:

[1327] Users enter keywords in the Market's search bar to find generative AI models optimized for a specific field (e.g., healthcare, finance, etc.).

[1328] Step 2:

[1329] The device sends the entered search query to the server, which includes the keywords entered by the user.

[1330] Step 3:

[1331] The server retrieves a list of relevant generative AI models from the database, and efficiently extracts relevant data using LIKE clauses and full-text search indexes.

[1332] Step 4:

[1333] The emotion engine analyzes the user's emotional data and prioritizes models that elicit positive emotional reactions. Emotional data is obtained by analyzing information collected from previous sessions, etc.

[1334] Step 5:

[1335] The server returns a list of models to the user in JSON format, which is then displayed on the user's device.

[1336] Step 6:

[1337] The user selects the model of interest from the returned list of models.

[1338] Step 7:

[1339] The device generates a request to send the selected model's ID to the server, which is also formatted as JSON.

[1340] Step 8:

[1341] The server retrieves detailed information about the model from the database. When retrieving, it uses a SELECT statement to extract detailed information about the model with the specified ID.

[1342] Step 9:

[1343] The server sends the detailed information it has obtained to the terminal in JSON format.

[1344] Step 10:

[1345] The device displays detailed information to the user, including a model description, performance metrics, price, and ratings.

[1346] Purchasing and using models

[1347] Step 1:

[1348] The user clicks the purchase button and enters payment information (such as credit card information). The entered payment information is encrypted to ensure security.

[1349] Step 2:

[1350] The device sends the purchase request and payment information to the server. The communication is encrypted using TLS / SSL during transmission.

[1351] Step 3:

[1352] The server then sends the payment information to an external payment processor via a secure channel using API integration.

[1353] Step 4:

[1354] The server receives the result of the payment process, and if successful, saves the success status along with the transaction ID.

[1355] Step 5:

[1356] The server issues a message confirming a successful purchase to the user along with access to the generative AI model, which is granted in the form of an additional session token.

[1357] Step 6:

[1358] Users access the download link or API endpoint to start using the purchased model.

[1359] Step 7:

[1360] The device sends a download request or API request to the server.

[1361] Step 8:

[1362] The server validates the user's session token to ensure access is permitted, and only if the validation is successful does it send the requested model file to the device.

[1363] Step 9:

[1364] The device unzips the received model file and begins using it in a local or cloud environment.

[1365] Model tuning and customization

[1366] Step 1:

[1367] If a user wishes to customize a purchased model, they open a customization request form within the system.

[1368] Step 2:

[1369] The terminal sends the customization request content (request details, specifications, dataset, etc.) entered by the user to the server.

[1370] Step 3:

[1371] The server forwards the received customization request to a specialized data scientist.

[1372] Step 4:

[1373] A data scientist receives the request and performs any necessary tuning. If any additional questions or clarifications are required, that information is also sent to the user via the server.

[1374] Step 5:

[1375] The server stores the completed tuning results in a database and notifies the user.

[1376] Step 6:

[1377] Users will be notified and will get new access information (download link, API endpoint, etc.) for the improved model.

[1378] Step 7:

[1379] The terminal sends an access request for the improved model to the server.

[1380] Step 8:

[1381] The server validates the user's session token to ensure access is permitted, and only if the validation is successful does it send the improved model file to the device.

[1382] Step 9:

[1383] The user unpacks the improved model they receive and starts using it in the required environment.

[1384] Monitoring and Maintenance

[1385] Step 1:

[1386] The server monitors the usage of the generated AI model in real time, including the frequency of API calls, error logs, and user operation history.

[1387] Step 2:

[1388] The server periodically collects and analyzes usage data, and if an anomaly is detected, an alert is generated and notifies the user.

[1389] Step 3:

[1390] Users receive notifications and can take action as needed, such as reporting a problem or suspending the model.

[1391] Step 4:

[1392] The server will check for any issues and take maintenance action as needed, including bug fixes and model retraining.

[1393] Step 5:

[1394] Once the server has completed its maintenance work, it will notify users that it is available again.

[1395] Incorporating an emotion engine

[1396] Step 1:

[1397] The server uses an emotion engine to collect emotional data in real time from the user's voice, text, facial expressions, etc. The collected data is temporarily stored in storage.

[1398] Step 2:

[1399] The server analyzes the collected emotion data and uses machine learning algorithms to determine the emotional state (e.g., stress, joy, sadness). This analysis can be done in batch or real-time.

[1400] Step 3:

[1401] Based on the emotion data acquired by the server, the system presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, models that are effective in reducing stress are prioritized.

[1402] Step 4:

[1403] The server's emotion engine periodically analyzes the emotion data and proposes or customizes models as needed. The analysis results are reflected in the next model proposal.

[1404] Step 5:

[1405] The user reviews the emotion-based suggested model and purchases or uses it. The suggested model is optimized for the user's current emotional state.

[1406] The above are the specific processing steps for implementing the present invention. The system provides domain-specific generative AI models, and can further improve the user experience by suggesting and customizing models based on the user's emotional state.

[1407] (Application example 2)

[1408] 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."

[1409] In the conventional AI model market, models based on user emotions were not provided, making it difficult to improve the user experience. Furthermore, it was difficult for users to find the optimal model under stressful circumstances, making it difficult to use the model efficiently. This resulted in low satisfaction with the purchase and use of AI models, limiting the overall effectiveness of the system.

[1410] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for building a marketplace that provides generative AI models specialized in specific fields, means for storing the generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and settling on the generative AI models, means for tuning the generative AI models in response to customization requests, means for monitoring the usage status of the generative AI models and performing necessary maintenance, means for recognizing the end user's emotions and proposing generative AI models based on the emotions, and means for analyzing emotion data and controlling the priority display of generative AI models. This makes it possible to provide personalized models according to the user's emotions, improving the user experience and dramatically improving the efficiency from model purchase to usage.

[1411] A "generative AI model" is an artificial intelligence model that is generated specifically for a specific field and exhibits high performance in specific tasks and data processing.

[1412] "Market" means an online marketplace provided for end users to search for, purchase, and use domain-specific generative AI models.

[1413] A "database" is a system for managing and storing data such as generative AI models, user information, transaction history, and emotional data.

[1414] "End User" means the ultimate user who accesses the system to utilize the Generative AI Model.

[1415] The "emotion engine" is a component that collects and analyzes emotional data in real time from end users' voices, text, facial expressions, etc.

[1416] A "Customization Request" is a request made by an End User to modify or adjust a Generative AI Model based on their specific needs or requirements.

[1417] "Tuning" is the process of adjusting and optimizing the performance and functionality of a generative AI model based on customization requirements.

[1418] "Priority display" is an operation that prominently presents generative AI models that are more appropriate and relevant to the user based on the results of emotional data analysis.

[1419] "Monitoring" is the process of monitoring the usage of generative AI models in real time and immediately detecting anomalies or problems.

[1420] "Maintenance" refers to the periodic management work of correcting and improving any problems or abnormalities that arise during use.

[1421] "Providing personalized models" means selecting and providing the optimal generative AI model based on the end user's individual emotional state and needs.

[1422] The present invention relates to a system for virtual stores that provides generative AI models specialized for specific fields. This system can recognize user emotions and propose optimal generative AI models based on those emotions. Specific embodiments for implementing this system are described below.

[1423] System configuration

[1424] 1. Building a Market:

[1425] The system will create a marketplace offering generative AI models in a virtual store, with a digital platform accessible to users that allows them to search, view details, purchase, and customize generative AI models.

[1426] 2. Use of databases:

[1427] The system uses a database to store and manage generative AI models, user information, transaction history, emotional data, etc. The database uses a common database management system such as MongoDB.

[1428] 3. User Device:

[1429] Users access the market using devices such as smartphones, tablets, and PCs, connecting to the system via a web browser or dedicated application.

[1430] 4. Server:

[1431] The server is the central component that receives user requests and performs the necessary processing. The server is built using a web framework such as Flask.

[1432] Specific processing steps

[1433] 1. User Registration and Authentication:

[1434] The server receives the information the user needs to create an account, validates it, and stores it in the database. When a user logs in, it checks the authentication information provided.

[1435] 2. Model exploration and selection:

[1436] The server receives a search query from the user's device and retrieves a list of relevant generative AI models from the database. The emotion engine analyzes the user's emotions and prioritizes generative AI models that evoke positive emotions.

[1437] 3. Use the Emotion Engine:

[1438] The server collects and analyzes emotional data from the user's touch points (e.g., microphone, camera) in real time. Based on the analysis results, it lists the generative AI model that is best suited to the user.

[1439] 4. Purchasing and using the model:

[1440] When a user purchases a generative AI model, the server processes the provided payment information. After purchase, the user can access the generative AI model and use it for download or via API.

[1441] 5. Processing customization requests:

[1442] The server forwards the customization request from the user to an expert (data scientist), who then provides a tuned generative AI model.

[1443] 6. Monitoring and Maintenance:

[1444] The server monitors the usage status of the generated AI model in real time and notifies the user if an abnormality is detected. Regular maintenance work is performed to maintain the suitability of the model for use.

[1445] Specific examples

[1446] 1. Example prompt:

[1447] "What field does the AI ​​model you're interested in relate to? Please be specific:"

[1448] "Are you feeling stressed? If so, would you suggest an AI model to help you relax?"

[1449] 2. Example of using the Emotion Engine:

[1450] When a user visits a virtual store, the emotion engine collects and analyzes the user's emotional data using a webcam and microphone. Based on the analysis results, the display priority of the model is controlled.

[1451] This invention makes it possible to provide an optimal generative AI model that corresponds to the user's emotions, thereby improving the user experience and enabling efficient model utilization.

[1452] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1453] Step 1: User Registration

[1454] Subject: User, Device, Server

[1455] Input: The user enters required information such as email address, password, and username on the account creation page.

[1456] Specific operation: The terminal sends the information entered by the user to the server as an HTTP request. The server validates the received information and, if there are no problems, saves the new user information in the database.

[1457] Output: The server returns a successful registration message to the user.

[1458] Step 2: User authentication

[1459] Subject: User, Device, Server

[1460] Input: The user enters their email address and password on the login page.

[1461] Specific operation: The terminal sends the entered authentication information to the server. The server searches the database for user information and performs authentication if there is a match.

[1462] Output: The server grants the user access to the market with a message of successful authentication.

[1463] Step 3: Model exploration

[1464] Subject: User, Device, Server

[1465] Input: A user enters a search query to explore generative AI models in a specific area within the market.

[1466] How it works: The device sends a search query to the server, which retrieves a list of relevant generative AI models from the database and analyzes the user's emotions using the emotion engine.

[1467] Output: The server returns a list of generated AI models that correspond to the user. Based on the emotion engine, models that match the user's emotions are displayed preferentially.

[1468] Step 4: Model selection

[1469] Subject: User, Device, Server

[1470] Input: The user selects the generative AI model of interest.

[1471] Specific operation: The device requests detailed information about the selected model. The server retrieves the details from the database and sends them to the device.

[1472] Output: The device displays detailed information about the generated AI model to the user.

[1473] Step 5: Purchase a model

[1474] Subject: User, Device, Server

[1475] Input: The user clicks the purchase button and enters payment information.

[1476] Specific operation: The device sends a purchase request and payment information to the server. The server processes the payment and, if successful, grants the user access to the generated AI model.

[1477] Output: The server sends the user a message confirming the purchase and providing an API key and download link.

[1478] Step 6: Use the model

[1479] Subject: User, Device, Server

[1480] Input: Uses a generative AI model purchased by the user.

[1481] Specific operation: The user accesses the generated AI model through their device. The server processes the API request and provides the required data and model to the user.

[1482] Output: The user uses the generative AI model for their own purposes.

[1483] Step 7: Customization Request

[1484] Subject: User, Device, Server

[1485] Input: The user details the customization requirements for the generative AI model.

[1486] How it works: The device sends a customization request to the server, which then forwards the request to a specialized data scientist.

[1487] Output: The server notifies the user when a new tuned model is available.

[1488] Step 8: Monitoring and Maintenance

[1489] Subject: Server

[1490] Input: Usage data for generative AI models

[1491] Specific operation: The server monitors the usage of the generative AI model in real time, analyzes the collected data, and if an abnormality is detected, notifies the user and performs necessary maintenance.

[1492] Output: The server evaluates the suitability of the model for use based on the results of the analysis of the usage data and makes corrections if necessary.

[1493] 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.

[1494] 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.

[1495] 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.

[1496] [Third embodiment]

[1497] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1498] 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.

[1499] 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).

[1500] 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.

[1501] 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.

[1502] 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).

[1503] 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.

[1504] 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.

[1505] 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.

[1506] 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.

[1507] 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.

[1508] 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."

[1509] A system for providing generative AI models specialized for specific fields is configured as follows.

[1510] System Overview

[1511] This system builds a marketplace that provides generative AI models specialized in specific fields, and allows users to access the marketplace to search, purchase, and use models. The system mainly consists of the following components:

[1512] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[1513] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes databases, authentication systems, and payment systems.

[1514] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, and transaction history.

[1515] Explanation of program processing

[1516] User Registration and Authentication

[1517] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[1518] Terminal: Sends the entered information to the server as an HTTP request.

[1519] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[1520] Server: Returns a successful registration message to the user.

[1521] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[1522] Model Exploration and Selection

[1523] Users: Explore generative AI models optimized for specific domains (e.g., healthcare, finance, etc.).

[1524] Device: Sends the user's search query to the server.

[1525] Server: Retrieves a list of relevant generative AI models from the database and returns them to the user.

[1526] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[1527] Server: Retrieves detailed information from the database and sends it to the device.

[1528] Terminal: Display detailed information to the user.

[1529] Purchasing and using models

[1530] User: Select the model they wish to purchase and enter their payment information to complete the purchase.

[1531] Terminal: Sends the entered payment information to the server.

[1532] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[1533] Users: Download purchased models or access them through API.

[1534] Model tuning and customization

[1535] User: If a purchased model needs customization, the user sends a request to the server with their specific requirements.

[1536] Server: Receives requests and relays them to expert data scientists.

[1537] Server: Data scientists tune the model and generate a new model.

[1538] Server: When a new model is ready, it notifies the user and provides access.

[1539] Monitoring and Maintenance

[1540] Server: Monitors the usage of the generative AI model in real time.

[1541] Server: Collects and analyzes usage data and notifies users if anomalies are detected, such as when a model responds excessively slowly or when the error rate is high.

[1542] Server: Once an abnormality is confirmed, maintenance work will be carried out and users will be notified that the service is available again once the problem has been fixed.

[1543] Specific examples

[1544] 1. User Registration and Authentication Example

[1545] Researchers at medical institutions register an account to use the system for the first time.

[1546] Researchers log in and search for generative AI models to find AI models specialized for pathological diagnosis.

[1547] 2. Example of model purchase

[1548] Researchers check detailed information about the pathology diagnostic AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[1549] After purchase, researchers can download the model and begin using it in their actual research.

[1550] 3. Model Tuning Example

[1551] A researcher requests customization, requesting detailed tuning based on clinical data.

[1552] Specialized data scientists will handle the process and provide newly tuned models.

[1553] This invention makes it possible to safely and effectively provide generative AI models specialized for specific fields, enabling users to expect high effectiveness in those fields. In addition, the ease of customization and maintenance improves user convenience.

[1554] The processing flow will be explained below.

[1555] User Registration and Authentication

[1556] Step 1:

[1557] User: Visit the Market and open the account creation page.

[1558] Step 2:

[1559] Device: Enter the required information (email address, password, username, etc.).

[1560] Step 3:

[1561] Terminal: Sends the entered information to the server as an HTTP request.

[1562] Step 4:

[1563] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[1564] Step 5:

[1565] Server: Returns a successful registration message to the user.

[1566] Step 6:

[1567] User: Opens the login page and enters their email address and password.

[1568] Step 7:

[1569] Terminal: Sends the entered authentication information to the server.

[1570] Step 8:

[1571] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[1572] Step 9:

[1573] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[1574] Model Exploration and Selection

[1575] Step 1:

[1576] User: Select a specific sector within the market (e.g., healthcare, finance, etc.).

[1577] Step 2:

[1578] On your device: Enter your search query in the search bar within Market.

[1579] Step 3:

[1580] Device: Sends a search query to the server.

[1581] Step 4:

[1582] Server: Obtain a list of relevant generative AI models from the database.

[1583] Step 5:

[1584] Server: Returns the retrieved model list to the user.

[1585] Step 6:

[1586] User: Select the model of interest from the returned list of models.

[1587] Step 7:

[1588] User: Sends a request to view detailed information about a selected model.

[1589] Step 8:

[1590] Server: Retrieves detailed information from the database and sends it to the device.

[1591] Step 9:

[1592] Terminal: Display detailed information to the user.

[1593] Purchasing and using models

[1594] Step 1:

[1595] User: Clicks the purchase button and enters payment information.

[1596] Step 2:

[1597] Terminal: Sends purchase request and payment information to the server.

[1598] Step 3:

[1599] Server: Processes the payment, and if successful, sends the user a purchase confirmation message and grants them access to the model.

[1600] Step 4:

[1601] Users: Download purchased models or access them via API.

[1602] Model tuning and customization

[1603] Step 1:

[1604] Users: If you require customization of the purchased model, please describe your request in detail.

[1605] Step 2:

[1606] Device: Sends a customization request to the server.

[1607] Step 3:

[1608] Server: Forwards the request to an expert data scientist.

[1609] Step 4:

[1610] Server: Data scientists perform tuning and create new generative AI models.

[1611] Step 5:

[1612] Server: Saves the new generative AI model in the database and notifies the user.

[1613] Step 6:

[1614] User: Use the improved model with the new access information.

[1615] Monitoring and Maintenance

[1616] Step 1:

[1617] Server: Monitors the usage of the generative AI model in real time.

[1618] Step 2:

[1619] Server: Collects and analyzes usage data.

[1620] Step 3:

[1621] Server: If an anomaly is detected, a notification is sent to the user.

[1622] Step 4:

[1623] Users: Receive notifications and take action as needed.

[1624] Step 5:

[1625] Server: If a problem is identified, maintenance work is carried out to fix it.

[1626] Step 6:

[1627] Server: Once the fix is ​​complete, notify users that it is available again.

[1628] In this way, the system provides generative AI models specialized for specific fields and can be effectively operated to meet the diverse needs of users.

[1629] Example 1

[1630] 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."

[1631] Platforms that provide generative AI models specialized in specific fields are required to build an environment where users can efficiently search, purchase, and use generative AI models. However, many current systems lack sufficient functionality for user authentication, search query analysis, customization request response, usage monitoring, and anomaly detection, making it difficult to improve user convenience and reliability.

[1632] 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.

[1633] In this invention, the server includes means for building a platform that provides generative AI models specialized in specific fields, means for storing generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for generative AI models, means for adjusting generative AI models in response to customization requests, means for monitoring the usage status of generative AI models and performing necessary maintenance work, means for users to log in and input authentication information, and means for analyzing user search queries and suggesting appropriate generative AI models. This enables users to efficiently search for, purchase, customize, and use generative AI models.

[1634] "Specific fields" refer to areas with specific expertise or needs, such as medicine, finance, or education.

[1635] "Generative AI models" refer to artificial intelligence models that generate new data and content using techniques such as generative adversarial networks (GANs) and natural language generation (NLG).

[1636] "Platform" refers to the infrastructure that provides an online environment for users to search for, purchase, and use generative AI models.

[1637] "Database" refers to a system for systematically storing and managing generative AI models, related user information, transaction history, etc.

[1638] "End User" refers to the ultimate user who utilizes the generative AI model to provide a specific application or service.

[1639] A "search query" refers to a keyword or phrase entered by a user when searching for a generative AI model.

[1640] "Customization Request" means a request submitted by a User to adjust or refine a Generative AI Model based on their specific needs.

[1641] "Maintenance work" refers to regular checks and maintenance work to ensure that generative AI models function properly.

[1642] "Login" refers to the authentication procedure required for a user to access a system.

[1643] "Authentication Information" refers to information such as email address and password that a User provides to prove access to a System.

[1644] "Search query analysis" refers to the process of analyzing a search query entered by a user and proposing the optimal generative AI model.

[1645] The system of the present invention builds a platform that provides generative AI models specialized in specific fields, and allows end users to search, purchase, customize, and use generative AI models using this platform. The main components of the system are as follows:

[1646] 1. User Device

[1647] A device that users access and operate. It can be a PC, smartphone, tablet, etc. Users access the system through their device to register an account, log in, search for models, view detailed information, make purchases, and request customization.

[1648] 2. Server

[1649] The server is the central component that manages and operates the entire system. It stores generative AI models in a database and includes an authentication system for user authentication, a search query analysis system, a payment system, and a system for responding to customization requests. It also provides API endpoints for users to access. The server communicates with the database and returns appropriate information based on the user's request.

[1650] 3. Database

[1651] This is a system for storing data such as generative AI models, user information, and transaction history. This database uses database systems such as PostgreSQL and Elasticsearch, particularly to speed up searches and maintain data consistency.

[1652] For example, a user accesses a market website and enters the required information (email address, password, username, etc.) on the account registration page. The device sends this information to the server, which validates it and stores it in the database. When the user enters their authentication information on the login page, the server authenticates the user and, if successful, allows them to access the market.

[1653] Next, the user searches for generative AI models suitable for a specific field (e.g., medicine or finance). The device sends the search query to the server, which retrieves and returns a list of relevant models from the database. The user selects the model of interest and requests that detailed information be displayed. The server retrieves the detailed information from the database and sends it to the device, allowing the user to purchase or customize the model.

[1654] Furthermore, if a user requests customization, the request is forwarded to the server and handled by a specialized data scientist. Once customization is complete, a new model is generated and provided to the user. The server also monitors the usage of the generated AI model in real time and takes appropriate action if an abnormality is detected.

[1655] Prompt Sentence Examples

[1656] 1. Researchers at medical institutions must register an account to use the system for the first time.

[1657] 2. Purchase and download the AI ​​model to be used for pathology diagnosis.

[1658] 3. Customize the pathology diagnostic AI model you purchased based on clinical data.

[1659] In this way, the system effectively provides domain-specific generative AI models, enhancing user convenience.

[1660] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1661] Step 1: User Registration

[1662] Users: Visit the Market registration page and enter your name, email address, and password.

[1663] Input: Name, email address, and password entered by the user into the form.

[1664] Output: Registration successful message.

[1665] Terminal: Sends user input information to the server in JSON format.

[1666] Input: Information entered by the user.

[1667] Output: HTTP POST request in JSON format.

[1668] Server: Validate the received information using a validation library (e.g. Joi) and, if there are no problems, save the user information to the PostgreSQL database.

[1669] Input: User information in JSON format.

[1670] Output: The new user information is saved in the database.

[1671] Server: Returns a successful registration message to the user in JSON format.

[1672] Input: User information saved successfully.

[1673] Output: Registration successful message.

[1674] Step 2: User Login

[1675] User: Access the login page and enter the registered email address and password.

[1676] Enter your email address and password.

[1677] Output: Login request.

[1678] Terminal: Sends the user's authentication information to the server in a JSON-formatted HTTP POST request.

[1679] Input: The credentials entered by the user.

[1680] Output: HTTP POST request in JSON format.

[1681] Server: Validates the received authentication information using a password authentication system (e.g. bcrypt) and issues a JWT token if it matches.

[1682] Input: User credentials.

[1683] Output: JWT token.

[1684] Server: Sends the JWT token to the user terminal.

[1685] Input: JWT token.

[1686] Output: User authentication success message and JWT token.

[1687] Step 3: Model Search

[1688] Users: Find generative AI models optimized for specific domains through search queries.

[1689] Input: Search query.

[1690] Output: The search request.

[1691] Terminal: Sends the user's query to the server as an HTTP GET request in JSON format.

[1692] Input: Search query.

[1693] Output: HTTP GET request in JSON format.

[1694] Server: Uses a search engine such as Elasticsearch to retrieve the relevant generative AI model from the database.

[1695] Input: Search query.

[1696] Output: A list of generative AI models.

[1697] Server: Returns search results to the user's device.

[1698] Input: A list of generative AI models.

[1699] Output: Search results.

[1700] Step 4: View model details

[1701] User: Select the model they are interested in and request more information.

[1702] Input: Model ID.

[1703] Output: Model details request.

[1704] Device: Send an HTTP GET request in JSON format containing the model ID to the server.

[1705] Input: Model ID.

[1706] Output: HTTP GET request in JSON format.

[1707] Server: Obtains detailed information about the selected generative AI model from the database and sends it to the user's device.

[1708] Input: Model ID.

[1709] Output: Detailed information about the model.

[1710] Terminal: Displays the received detailed information to the user.

[1711] Input: Model details.

[1712] Output: The detailed information displayed to the user.

[1713] Step 5: Purchase a model

[1714] User: Selects the model they want to purchase and enters their payment information to begin the checkout process.

[1715] Input: Payment information.

[1716] Output: Purchase request.

[1717] Terminal: Sends the user's payment information to the server via an HTTPS POST request in JSON format.

[1718] Input: Payment information.

[1719] Output: HTTPS POST request in JSON format.

[1720] Server: Process the payment using the Stripe API, and if successful, record the purchase confirmation message and model access rights in the database.

[1721] Input: Payment information.

[1722] Output: Purchase successful message and updated access rights.

[1723] Server: Sends a purchase confirmation message and access right information to the user terminal.

[1724] Input: Purchase success message and access rights information.

[1725] Output: Notification to user terminal.

[1726] Step 6: Use the model

[1727] Users: Access or download purchased models via API.

[1728] Input: Access request.

[1729] Output: Model use.

[1730] Server: Receives requests to access a model, performs appropriate authentication, and returns the model file or issues an API key.

[1731] Input: Access request.

[1732] Output: Model file or API key.

[1733] Terminal: Provide the received model file or API key to the user.

[1734] Input: Model file or API key.

[1735] Output: The model is made available to the user.

[1736] (Application example 1)

[1737] 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."

[1738] Modern factories require efficient and rapid optimization of manufacturing processes. It is difficult to easily obtain and use generative AI models suited to specific manufacturing environments. Furthermore, customization of generative AI models and effective monitoring after their deployment are required. To address these challenges, an effective means is needed to easily provide AI models for optimizing the operation of factory robots, and to customize and maintain those models.

[1739] 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.

[1740] In this invention, the server includes means for building a marketplace that provides generative AI models specialized in specific fields, means for storing the generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for the generative AI models, means for tuning the generative AI models in response to customization requests, means for monitoring the usage status of the generative AI models and performing necessary maintenance, means for end users to select, download, or integrate generative AI models to be applied to factory robots, and means for optimizing the operation of the factory robots for specific manufacturing processes. This enables end users to easily obtain generative AI models suitable for their manufacturing environment and effectively apply and operate them.

[1741] A "generative AI model" is an algorithmic model of artificial intelligence that is generated for a specific field or application, and exhibits performance optimized for that field.

[1742] "Market" means the online platform where end users can find, purchase, customize, download, and integrate Generative AI Models.

[1743] A "database" is an information recording system for managing and storing generative AI models, user information, transaction history, etc.

[1744] "End users" are the final users of generative AI models, such as factory workers and managers.

[1745] "Customization Request" means a request by an End User to tailor a Generative AI Model based on their specific needs and requirements.

[1746] "Tuning" refers to the process of adjusting a generative AI model to optimize it for the end user's specific use and environment.

[1747] "Monitoring" refers to the act of monitoring the usage and performance of generative AI models in real time.

[1748] "Maintenance" refers to maintenance work to maintain the performance and reliability of the generative AI model, including fixing bugs and making improvements.

[1749] A "factory robot" is a mechanical device used to automate specific tasks in manufacturing sites, and its operations are controlled by a program.

[1750] "Manufacturing process" refers to the series of operations or steps that transform raw materials into a finished product, including welding, assembly, quality inspection, etc.

[1751] "Integration" refers to the act of end users applying purchased generative AI models to their own systems or machinery (e.g., factory robots) to make them work together.

[1752] This invention relates to a system that enables end users to easily search, purchase, customize, download, and integrate generative AI models suited to specific manufacturing processes. This system is primarily composed of the following hardware and software:

[1753] Hardware and Software Use Cases

[1754] Hardware: A factory robot (e.g., a generic robotic arm) and the smartphone or tablet (iOS or Android device) that controls it.

[1755] software:

[1756] Robot control software (e.g., Robot Operating System (ROS)).

[1757] A cloud service (e.g., AWS or Google Cloud) as a model management server.

[1758] The application for operation on smart devices will be developed using React Native.

[1759] Program processing explanation

[1760] In this invention, the server, terminal, and user work together to search, select, purchase, customize, and use generative AI models. The specific processing steps are as follows:

[1761] 1. User Registration and Authentication

[1762] Users open the app on their smartphone and enter the required information on the account creation page.

[1763] The terminal transmits the input information to the server as an HTTP request.

[1764] The server validates the received information and, if there are no problems, saves the new user information in the database and returns a message to the user confirming successful registration.

[1765] 2. Model Search and Selection

[1766] Users enter a search query to explore generative AI models specific to their manufacturing process.

[1767] The terminal transmits the user's search query to the server.

[1768] The server retrieves a list of relevant generative AI models from the database and returns it to the user.

[1769] The user selects a model of interest from the returned list and requests that its detailed information be displayed.

[1770] The server retrieves the detailed information from the database and sends it to the terminal.

[1771] The terminal displays detailed information to the user.

[1772] 3. Purchasing and Using the Model

[1773] Users select the generative AI model they want to purchase and enter their payment information.

[1774] The terminal transmits the entered payment information to the server.

[1775] The server processes the payment and, if successful, gives the user a purchase confirmation message and access to the model.

[1776] Users can download the purchased model or integrate it into their robot control system via API.

[1777] 4. Tuning and customizing the model

[1778] When a user desires tuning specialized for a specific production line, the user writes a customization request and sends it to the server.

[1779] The server receives the request and relays it to the data scientist.

[1780] The server will notify the user and grant access as soon as a new tuned model is available.

[1781] 5. Monitoring and Maintenance

[1782] The server monitors the usage of the generative AI model in real time.

[1783] The server collects and analyzes usage data and notifies the user if an anomaly is detected, such as a robot slowing down or an increased error rate.

[1784] The server will perform maintenance when an abnormality is detected and notify users that the service is available again once the problem has been fixed.

[1785] Specific examples

[1786] For example, if an end user wants to apply an AI model to a welding process in a factory, they might enter a prompt like this:

[1787] "Optimize the welding seam angle and speed used by this robotic arm."

[1788] This allows the generative AI model to optimize the parameters of the welding process, reducing error rates and improving production speed.

[1789] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1790] Step 1: User Registration and Authentication

[1791] User: Opens the smartphone app and enters the required information (email address, password, factory ID, etc.) on the account creation page.

[1792] Terminal: Sends the entered information to the server as an HTTP request.

[1793] Server: Validates the received information and, if there are no problems, saves the new user information to the database. If the save is successful, generates a message indicating successful registration and returns it to the terminal. Examples of validation include detecting invalid email addresses and passwords that are too short.

[1794] Step 2: Log in

[1795] User: Enters email address and password on the login page and clicks the login button.

[1796] Terminal: Sends the entered information to the server as an HTTP request.

[1797] Server: Searches for the corresponding user information in the database and compares it with the authentication information. If it matches, it generates a login success message and session information and sends them to the terminal.

[1798] Step 3: Model exploration

[1799] User: Enter a specific manufacturing process (e.g., welding, assembly) into the app's search bar and press the search button.

[1800] Terminal: Sends the search query as an HTTP request to the server.

[1801] Server: Searches the database for a list of relevant generative AI models, generates a list of relevant models, and sends it to the device. Specifically, it filters the models by tags and keywords related to the manufacturing process.

[1802] Step 4: Model selection and detailed display

[1803] User: Selects the generative AI model of interest from the returned list and requests more information about it.

[1804] Device: Send the model ID to the server as an HTTP request.

[1805] Server: Retrieves detailed information about the selected model from the database, generates detailed information, and sends it to the terminal. The detailed information includes the model's scope of application, expected results, and usage methods.

[1806] Step 5: Purchase and pay for the model

[1807] User: Select the generative AI model they want to purchase and enter their payment information (credit card number, security code, etc.).

[1808] Terminal: Sends payment information to the server as an HTTP request.

[1809] Server: Works with the payment processing system to perform payment, and if successful, generates a purchase confirmation message and access rights for the model and sends them to the terminal. If unsuccessful, generates an error message and sends it to the terminal.

[1810] Step 6: Download and integrate the model

[1811] User: Presses a button to download the purchased generative AI model.

[1812] Device: Sends a model download request to the server.

[1813] Server: Retrieves the corresponding model file from the database, generates a download link, and sends it to the device.

[1814] Users: Click the download link and integrate the model file into their robot control system. Specifically, they load the model using software such as ROS and set parameters to optimize the robot's behavior.

[1815] Step 7: Tune and customize the model

[1816] User: If a user wants specialized tuning for a specific production line, he or she enters a customization request and sends it to the server.

[1817] Terminal: Sends the customization request to the server as an HTTP request.

[1818] Server: Receives requests and forwards them to data scientists, who tune the model based on the requests and generate new versions of the model.

[1819] Server: Saves the new tuned model to the database and notifies the user of the update.

[1820] Step 8: Monitoring and Maintenance

[1821] Server: Collects logging and performance metrics to monitor the usage of generative AI models in real time.

[1822] Server: If abnormal behavior (e.g., delayed response, increased error rate) is detected, the server identifies the anomaly based on an anomaly detection algorithm and notifies the user. Statistical anomaly detection techniques and machine learning algorithms are used for anomaly detection.

[1823] Server: If an anomaly is detected, the maintenance team is notified and the model is available again once the fix is ​​complete. Continuous monitoring ensures the reliability and effectiveness of the model.

[1824] 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.

[1825] The system for providing generative AI models specialized for specific fields is configured as follows: This system also incorporates an emotion engine that recognizes user emotions.

[1826] System Overview

[1827] This system creates a marketplace that provides generative AI models specialized in specific fields, allowing users to access the marketplace and search, purchase, and use models. It also has the ability to recognize user emotions and suggest models and interact with them based on those emotions.

[1828] The system mainly consists of the following components:

[1829] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[1830] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes a database, authentication system, payment system, emotion engine, etc.

[1831] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, transaction history, and emotional data.

[1832] 4. Emotion Engine: A component for recognizing and analyzing user emotions.

[1833] Explanation of program processing

[1834] User Registration and Authentication

[1835] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[1836] Terminal: Sends the entered information to the server as an HTTP request.

[1837] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[1838] Server: Returns a successful registration message to the user.

[1839] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[1840] Model Exploration and Selection

[1841] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[1842] Device: Sends the user's search query to the server.

[1843] Server: Retrieves a list of applicable generative AI models from the database and returns it to the user. The emotion engine recognizes the user's emotions and prioritizes models that evoke positive emotional responses.

[1844] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[1845] Server: Retrieves detailed information from the database and sends it to the device.

[1846] Terminal: Display detailed information to the user.

[1847] Purchasing and using models

[1848] User: Clicks the purchase button and enters payment information.

[1849] Terminal: Sends purchase request and payment information to the server.

[1850] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[1851] Users: Download purchased models or access them via API.

[1852] Model tuning and customization

[1853] Users: If you require customization of the purchased model, please describe your request in detail.

[1854] Device: Sends a customization request to the server.

[1855] Server: Forwards the request to an expert data scientist.

[1856] Server: Data scientists perform tuning and create new generative AI models.

[1857] Server: Saves the new generative AI model in the database and notifies the user.

[1858] User: Use the improved model with the new access information.

[1859] Monitoring and Maintenance

[1860] Server: Monitors the usage of the generative AI model in real time.

[1861] Server: Collects and analyzes usage data.

[1862] Server: If an anomaly is detected, a notification is sent to the user.

[1863] Users: Receive notifications and take action as needed.

[1864] Server: If a problem is identified, maintenance work is carried out to fix it.

[1865] Server: Once the fix is ​​complete, notify users that it is available again.

[1866] Incorporating an emotion engine

[1867] Server: Builds an emotion engine and collects emotion data in real time from users' voices, text, facial expressions, etc.

[1868] Server: The emotion engine analyzes the collected data and determines the emotional state.

[1869] Server: Based on the acquired emotion data, the server presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, the server prioritizes models that are effective in reducing stress.

[1870] Server: The emotion engine periodically analyzes the emotion data and suggests or customizes models as needed.

[1871] User: Review the sentiment-based recommendation model and make a purchase or use it.

[1872] Specific examples

[1873] 1. User Registration and Authentication Example

[1874] A pharmaceutical researcher from a healthcare company registers an account to use the system for the first time.

[1875] Researchers log in and search generative AI models to find AI models specialized for specific diseases.

[1876] 2. Example of model purchase

[1877] Researchers check detailed information about the disease-specific AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[1878] After purchase, researchers can download the model and begin using it in their research.

[1879] 3. Model Tuning Example

[1880] A researcher requests customization, requesting detailed tuning based on clinical data.

[1881] Specialized data scientists will handle the process and provide newly tuned models.

[1882] 4. Example of an Emotion Engine

[1883] As researchers use the system, the emotion engine analyzes their context and recognizes when they are feeling stressed.

[1884] Based on the emotion engine, we present effective guidelines for stress reduction and propose a model to improve researchers' work efficiency.

[1885] This invention enables the safe and effective provision of generative AI models specialized for specific fields. In addition, by incorporating an emotion engine, it is possible to provide more personalized services according to the user's emotional state.

[1886] The processing flow will be explained below.

[1887] User Registration and Authentication

[1888] Step 1:

[1889] User: Visit the Market and open the account creation page.

[1890] Step 2:

[1891] Device: Enter the required information (email address, password, username, etc.).

[1892] Step 3:

[1893] Terminal: Sends the entered information to the server as an HTTP request.

[1894] Step 4:

[1895] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[1896] Step 5:

[1897] Server: Returns a successful registration message to the user.

[1898] Step 6:

[1899] User: Opens the login page and enters their email address and password.

[1900] Step 7:

[1901] Terminal: Sends the entered authentication information to the server.

[1902] Step 8:

[1903] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[1904] Step 9:

[1905] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[1906] Model Exploration and Selection

[1907] Step 1:

[1908] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[1909] Step 2:

[1910] On your device: Enter your search query in the search bar within Market.

[1911] Step 3:

[1912] Device: Sends a search query to the server.

[1913] Step 4:

[1914] Server: Obtain a list of relevant generative AI models from the database.

[1915] Step 5:

[1916] Server: Returns the acquired model list to the user device.

[1917] Step 6:

[1918] User: Select the model of interest from the returned list of models.

[1919] Step 7:

[1920] User: Sends a request to view detailed information about a selected model.

[1921] Step 8:

[1922] Server: Retrieves detailed information from the database and sends it to the device.

[1923] Step 9:

[1924] Terminal: Display detailed information to the user.

[1925] Step 10:

[1926] Emotion engine: Analyzes the user's facial expressions, voice, text, etc. to obtain emotional data.

[1927] Step 11:

[1928] Server: Based on the emotion data, add the generative AI model that is suitable for the user to the suggestion list.

[1929] Purchasing and using models

[1930] Step 1:

[1931] User: Clicks the purchase button and enters payment information.

[1932] Step 2:

[1933] Terminal: Sends purchase request and payment information to the server.

[1934] Step 3:

[1935] Server: Processes the payment, and if successful, sends the user a purchase confirmation message and grants them access to the model.

[1936] Step 4:

[1937] Users: Download purchased models or access them through the API.

[1938] Model tuning and customization

[1939] Step 1:

[1940] Users: If you need customization for the model you purchased, please submit your specific request.

[1941] Step 2:

[1942] Terminal: Forwards the customization request to the server.

[1943] Step 3:

[1944] Server: Forwards the request to an expert data scientist.

[1945] Step 4:

[1946] Server: Data scientists tune the model and generate a new generative AI model.

[1947] Step 5:

[1948] Server: Saves the newly tuned generative AI model in the database and notifies the user.

[1949] Step 6:

[1950] User: Use the improved model with the new access information.

[1951] Monitoring and Maintenance

[1952] Step 1:

[1953] Server: Monitors the usage of the generative AI model in real time.

[1954] Step 2:

[1955] Server: Collects and analyzes usage data.

[1956] Step 3:

[1957] Server: If an anomaly is detected, a notification is sent to the user.

[1958] Step 4:

[1959] Users: Receive notifications and take action as needed.

[1960] Step 5:

[1961] Server: If a problem is identified, maintenance work is carried out to fix it.

[1962] Step 6:

[1963] Server: Once the fix is ​​complete, notify users that it is available again.

[1964] Incorporating an emotion engine

[1965] Step 1:

[1966] Server: Builds an emotion engine and collects emotion data in real time from users' voices, text, facial expressions, etc.

[1967] Step 2:

[1968] Server: The emotion engine analyzes the collected data and determines the emotional state.

[1969] Step 3:

[1970] Server: Based on the acquired emotion data, presents the user with a list of optimal generative AI models.

[1971] Step 4:

[1972] Server: Adjusts suggestions and interactions to the user based on emotion data. For example, if the user is feeling stressed, it makes suggestions to help them relax.

[1973] Specific examples

[1974] User registration and authentication examples

[1975] A researcher opens the account registration page.

[1976] The researcher enters the required information and completes the registration.

[1977] The researcher logs in using their email address and password.

[1978] Specific examples of model purchases

[1979] Researchers select a disease-specific AI model and view detailed information.

[1980] The researcher decides to purchase and enters payment information.

[1981] Researchers can download the model and use it in their research.

[1982] Example of model tuning

[1983] The researcher details and submits the customization request.

[1984] The server forwards the request to an expert data scientist.

[1985] Data scientists tune the model and provide a new model.

[1986] Examples of emotion engines

[1987] While researchers are using the system, an emotion engine detects their stress levels.

[1988] The server proposes a model suitable for stress reduction.

[1989] Researchers will follow the suggestions and use the model to improve operational efficiency.

[1990] This invention makes it possible to provide generative AI models specialized for specific fields while providing individual services according to the user's emotional state. The incorporation of an emotion engine is expected to improve the user experience and lead to more effective use.

[1991] Example 2

[1992] 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."

[1993] In recent years, the use of generative AI models has expanded in various fields, but there is a lack of systems that efficiently provide generative AI models that can meet specialized requirements.In addition, the lack of models that adapt to the user's emotional state often leads to a decrease in user satisfaction and hinders the effective use of the models.

[1994] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for building a market that provides generative AI models specialized in a specific field, a means for storing the generative AI models in a database, and a means for analyzing user emotion data and proposing an optimal generative AI model based on the analyzed data. This makes it possible to efficiently provide specialized generative AI models and propose and use models that are adapted to the user's emotional state.

[1995] "Market" is an online platform for providing domain-specific generative AI models.

[1996] A "generative AI model" is an artificial intelligence model that is generated based on data from a specific field and automates or assists with various tasks in that field.

[1997] A "database" is a system for storing generative AI models, user information, emotional data, etc.

[1998] "End User" means an individual or legal entity that uses the System to search for, purchase, and use Generative AI Models.

[1999] "Search Means" means means that provide the ability for an End User to search for a particular Generative AI Model.

[2000] "Means for displaying detailed information" refers to means for providing a function for end users to display detailed information (e.g., performance indicators, price, usage methods, etc.) about the generated AI model in which they are interested.

[2001] "Means for purchasing and payment" refers to means that provide the functionality for processing payments when end users purchase generative AI models.

[2002] A "customization request" is a request by an end user for additional functionality or tuning of a generative AI model that they have purchased.

[2003] "Means for tuning" means means for providing the ability to adjust or improve a generative AI model in response to end-user customization requests.

[2004] "Monitoring means" refers to a means for monitoring the usage of the generated AI model in real time and providing the functionality to detect anomalies and collect data.

[2005] "Maintenance measures" are measures for carrying out corrective work or maintenance to resolve abnormalities or problems detected through monitoring.

[2006] "Emotional data" is data that indicates the user's emotional state and is collected from voice, text, facial expressions, etc.

[2007] The "means for analyzing emotional data" is a means for analyzing collected emotional data and providing a function for determining the emotional state of the user.

[2008] The "emotion engine" is a system component that proposes and adjusts models based on the user's emotional state.

[2009] The present invention relates to a system that provides generative AI models specialized in specific fields. The system allows users to access, search, purchase, and use specialized generative AI models, and further has the function of analyzing user emotion data and proposing optimal generative AI models based on the data.

[2010] System Overview

[2011] The system consists of the following components:

[2012] 1. User terminal: A device that a user accesses and operates. Examples include PCs, smartphones, tablets, etc.

[2013] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. The server includes a database, authentication system, payment system, emotion engine, etc.

[2014] 3. Database: Stores data such as domain-specific generative AI models, user information, transaction history, and sentiment data.

[2015] 4. Emotion Engine: A component for recognizing and analyzing user emotions.

[2016] User Registration and Authentication

[2017] When a user accesses the market, they first enter the required information on the account creation page. The device sends this information to the server as an HTTP request. The server validates the received information and, if there are no problems, saves the new user information in the database. The user then enters their authentication information on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[2018] Model Exploration and Selection

[2019] When a user searches the market for a generative AI model optimized for a specific field, the device sends a search query to the server. The server retrieves a list of relevant generative AI models from the database and returns it to the user. The emotion engine recognizes the user's emotions and prioritizes models that evoke positive emotional reactions. The user selects a model of interest from the returned list of models and requests that its detailed information be displayed on the device. The server retrieves the detailed information from the database and sends it to the device. The device displays this information to the user.

[2020] Purchasing and using models

[2021] When a user purchases a model, they click the purchase button and enter their payment information. The device sends this request and payment information to the server. The server processes the payment through an external payment gateway, and if successful, issues the user a purchase confirmation message and access rights to the model. The user can then download the purchased model or access it via API.

[2022] Model tuning and customization

[2023] If a user wants to customize a model, they send a detailed request from their device to the server. The server then forwards the request to a specialized data scientist. Once the data scientist performs tuning and a new generative AI model is created, the server stores it in a database and notifies the user. The user can then use the improved model with the new access information.

[2024] Monitoring and Maintenance

[2025] The server monitors the usage of the generative AI model in real time, collecting and analyzing usage data. If an abnormality is detected, a notification is sent to the user. The user receives the notification and can take action as necessary. If a problem is confirmed, the server performs maintenance work, and once the problem is fixed, the server notifies the user that the model is available again.

[2026] Incorporating an emotion engine

[2027] The emotion engine allows the server to collect and analyze emotional data from the user's voice, text, facial expressions, etc. in real time. Based on the acquired emotional data, the server presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, models that are effective in reducing stress will be prioritized. The emotion engine regularly analyzes the emotional data and suggests or customizes models as needed.

[2028] Specific examples

[2029] 1. User Registration and Authentication Example

[2030] A pharmaceutical researcher from a healthcare company registers an account to use the system for the first time.

[2031] Researchers log in and search generative AI models to find AI models specialized for specific diseases.

[2032] 2. Example of model purchase

[2033] Researchers check detailed information about the disease-specific AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[2034] After purchase, researchers can download the model and begin using it in their research.

[2035] 3. Model Tuning Example

[2036] A researcher requests customization, requesting detailed tuning based on clinical data.

[2037] Specialized data scientists will handle the process and provide newly tuned models.

[2038] 4. Example of an Emotion Engine

[2039] As researchers use the system, the emotion engine analyzes their context and recognizes when they are feeling stressed.

[2040] Based on the emotion engine, we present effective guidelines for stress reduction and propose a model to improve researchers' work efficiency.

[2041] As a result, the present invention can safely and effectively provide generative AI models specialized for specific fields, while also providing personalized services that correspond to the user's emotional state.

[2042] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2043] User Registration and Authentication

[2044] Step 1:

[2045] The user accesses the market and enters the required information (email address, password, username, etc.) on the account creation page. The entered information is displayed on the user's device, and once confirmation is complete, the user presses the send button.

[2046] Step 2:

[2047] The device sends the entered information to the server as an HTTP request, which contains the entered information and is formatted in JSON.

[2048] Step 3:

[2049] The server validates the information it receives. For example, it checks the format of the email address or the strength of the password. This validation is performed using regular expressions or rule-based checks. If the input is found to be valid, it proceeds.

[2050] Step 4:

[2051] The server stores the successfully validated information in the database. The password is hashed and stored securely along with the user information. The storage process uses an SQL query.

[2052] Step 5:

[2053] The server returns a message to the user confirming successful registration. The message is structured in JSON format and sent to the terminal for display.

[2054] Step 6:

[2055] The user again enters their authentication information (email address, password) on the login page. The entered information is again sent by the terminal to the server as an HTTP request.

[2056] Step 7:

[2057] The server checks the received authentication information against its database, specifically to ensure that the hashed password matches. If the check is successful, it issues a session token to the user and returns a response containing that session token.

[2058] Step 8:

[2059] Users can access the market using a session token, which is stored in a browser cookie and used to simplify authentication in future visits.

[2060] Model Exploration and Selection

[2061] Step 1:

[2062] Users enter keywords in the Market's search bar to find generative AI models optimized for a specific field (e.g., healthcare, finance, etc.).

[2063] Step 2:

[2064] The device sends the entered search query to the server, which includes the keywords entered by the user.

[2065] Step 3:

[2066] The server retrieves a list of relevant generative AI models from the database, and efficiently extracts relevant data using LIKE clauses and full-text search indexes.

[2067] Step 4:

[2068] The emotion engine analyzes the user's emotional data and prioritizes models that elicit positive emotional reactions. Emotional data is obtained by analyzing information collected from previous sessions, etc.

[2069] Step 5:

[2070] The server returns a list of models to the user in JSON format, which is then displayed on the user's device.

[2071] Step 6:

[2072] The user selects the model of interest from the returned list of models.

[2073] Step 7:

[2074] The device generates a request to send the selected model's ID to the server, which is also formatted as JSON.

[2075] Step 8:

[2076] The server retrieves detailed information about the model from the database. When retrieving, it uses a SELECT statement to extract detailed information about the model with the specified ID.

[2077] Step 9:

[2078] The server sends the detailed information it has obtained to the terminal in JSON format.

[2079] Step 10:

[2080] The device displays detailed information to the user, including a model description, performance metrics, price, and ratings.

[2081] Purchasing and using models

[2082] Step 1:

[2083] The user clicks the purchase button and enters payment information (such as credit card information). The entered payment information is encrypted to ensure security.

[2084] Step 2:

[2085] The device sends the purchase request and payment information to the server. The communication is encrypted using TLS / SSL during transmission.

[2086] Step 3:

[2087] The server then sends the payment information to an external payment processor via a secure channel using API integration.

[2088] Step 4:

[2089] The server receives the result of the payment process, and if successful, saves the success status along with the transaction ID.

[2090] Step 5:

[2091] The server issues a message confirming a successful purchase to the user along with access to the generative AI model, which is granted in the form of an additional session token.

[2092] Step 6:

[2093] Users access the download link or API endpoint to start using the purchased model.

[2094] Step 7:

[2095] The device sends a download request or API request to the server.

[2096] Step 8:

[2097] The server validates the user's session token to ensure access is permitted, and only if the validation is successful does it send the requested model file to the device.

[2098] Step 9:

[2099] The device unzips the received model file and begins using it in a local or cloud environment.

[2100] Model tuning and customization

[2101] Step 1:

[2102] If a user wishes to customize a purchased model, they open a customization request form within the system.

[2103] Step 2:

[2104] The terminal sends the customization request content (request details, specifications, dataset, etc.) entered by the user to the server.

[2105] Step 3:

[2106] The server forwards the received customization request to a specialized data scientist.

[2107] Step 4:

[2108] A data scientist receives the request and performs any necessary tuning. If any additional questions or clarifications are required, that information is also sent to the user via the server.

[2109] Step 5:

[2110] The server stores the completed tuning results in a database and notifies the user.

[2111] Step 6:

[2112] Users will be notified and will get new access information (download link, API endpoint, etc.) for the improved model.

[2113] Step 7:

[2114] The terminal sends an access request for the improved model to the server.

[2115] Step 8:

[2116] The server validates the user's session token to ensure access is permitted, and only if the validation is successful does it send the improved model file to the device.

[2117] Step 9:

[2118] The user unpacks the improved model they receive and starts using it in the required environment.

[2119] Monitoring and Maintenance

[2120] Step 1:

[2121] The server monitors the usage of the generated AI model in real time, including the frequency of API calls, error logs, and user operation history.

[2122] Step 2:

[2123] The server periodically collects and analyzes usage data, and if an anomaly is detected, an alert is generated and notifies the user.

[2124] Step 3:

[2125] Users receive notifications and can take action as needed, such as reporting a problem or suspending the model.

[2126] Step 4:

[2127] The server will check for any issues and take maintenance action as needed, including bug fixes and model retraining.

[2128] Step 5:

[2129] Once the server has completed its maintenance work, it will notify users that it is available again.

[2130] Incorporating an emotion engine

[2131] Step 1:

[2132] The server uses an emotion engine to collect emotional data in real time from the user's voice, text, facial expressions, etc. The collected data is temporarily stored in storage.

[2133] Step 2:

[2134] The server analyzes the collected emotion data and uses machine learning algorithms to determine the emotional state (e.g., stress, joy, sadness). This analysis can be done in batch or real-time.

[2135] Step 3:

[2136] Based on the emotion data acquired by the server, the system presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, models that are effective in reducing stress are prioritized.

[2137] Step 4:

[2138] The server's emotion engine periodically analyzes the emotion data and proposes or customizes models as needed. The analysis results are reflected in the next model proposal.

[2139] Step 5:

[2140] The user reviews the emotion-based suggested model and purchases or uses it. The suggested model is optimized for the user's current emotional state.

[2141] The above are the specific processing steps for implementing the present invention. The system provides domain-specific generative AI models, and can further improve the user experience by suggesting and customizing models based on the user's emotional state.

[2142] (Application example 2)

[2143] 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."

[2144] In the conventional AI model market, models based on user emotions were not provided, making it difficult to improve the user experience. Furthermore, it was difficult for users to find the optimal model under stressful circumstances, making it difficult to use the model efficiently. This resulted in low satisfaction with the purchase and use of AI models, limiting the overall effectiveness of the system.

[2145] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for building a marketplace that provides generative AI models specialized in specific fields, means for storing the generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and settling on the generative AI models, means for tuning the generative AI models in response to customization requests, means for monitoring the usage status of the generative AI models and performing necessary maintenance, means for recognizing the end user's emotions and proposing generative AI models based on the emotions, and means for analyzing emotion data and controlling the priority display of generative AI models. This makes it possible to provide personalized models according to the user's emotions, improving the user experience and dramatically improving the efficiency from model purchase to usage.

[2146] A "generative AI model" is an artificial intelligence model that is generated specifically for a specific field and exhibits high performance in specific tasks and data processing.

[2147] "Market" means an online marketplace provided for end users to search for, purchase, and use domain-specific generative AI models.

[2148] A "database" is a system for managing and storing data such as generative AI models, user information, transaction history, and emotional data.

[2149] "End User" means the ultimate user who accesses the system to utilize the Generative AI Model.

[2150] The "emotion engine" is a component that collects and analyzes emotional data in real time from end users' voices, text, facial expressions, etc.

[2151] A "Customization Request" is a request made by an End User to modify or adjust a Generative AI Model based on their specific needs or requirements.

[2152] "Tuning" is the process of adjusting and optimizing the performance and functionality of a generative AI model based on customization requirements.

[2153] "Priority display" is an operation that prominently presents generative AI models that are more appropriate and relevant to the user based on the results of emotional data analysis.

[2154] "Monitoring" is the process of monitoring the usage of generative AI models in real time and immediately detecting anomalies or problems.

[2155] "Maintenance" refers to the periodic management work of correcting and improving any problems or abnormalities that arise during use.

[2156] "Providing personalized models" means selecting and providing the optimal generative AI model based on the end user's individual emotional state and needs.

[2157] The present invention relates to a system for virtual stores that provides generative AI models specialized for specific fields. This system can recognize user emotions and propose optimal generative AI models based on those emotions. Specific embodiments for implementing this system are described below.

[2158] System configuration

[2159] 1. Building a Market:

[2160] The system will create a marketplace offering generative AI models in a virtual store, with a digital platform accessible to users that allows them to search, view details, purchase, and customize generative AI models.

[2161] 2. Use of databases:

[2162] The system uses a database to store and manage generative AI models, user information, transaction history, emotional data, etc. The database uses a common database management system such as MongoDB.

[2163] 3. User Device:

[2164] Users access the market using devices such as smartphones, tablets, and PCs, connecting to the system via a web browser or dedicated application.

[2165] 4. Server:

[2166] The server is the central component that receives user requests and performs the necessary processing. The server is built using a web framework such as Flask.

[2167] Specific processing steps

[2168] 1. User Registration and Authentication:

[2169] The server receives the information the user needs to create an account, validates it, and stores it in the database. When a user logs in, it checks the authentication information provided.

[2170] 2. Model exploration and selection:

[2171] The server receives a search query from the user's device and retrieves a list of relevant generative AI models from the database. The emotion engine analyzes the user's emotions and prioritizes generative AI models that evoke positive emotions.

[2172] 3. Use the Emotion Engine:

[2173] The server collects and analyzes emotional data from the user's touch points (e.g., microphone, camera) in real time. Based on the analysis results, it lists the generative AI model that is best suited to the user.

[2174] 4. Purchasing and using the model:

[2175] When a user purchases a generative AI model, the server processes the provided payment information. After purchase, the user can access the generative AI model and use it for download or via API.

[2176] 5. Processing customization requests:

[2177] The server forwards the customization request from the user to an expert (data scientist), who then provides a tuned generative AI model.

[2178] 6. Monitoring and Maintenance:

[2179] The server monitors the usage status of the generated AI model in real time and notifies the user if an abnormality is detected. Regular maintenance work is performed to maintain the suitability of the model for use.

[2180] Specific examples

[2181] 1. Example prompt:

[2182] "What field does the AI ​​model you're interested in relate to? Please be specific:"

[2183] "Are you feeling stressed? If so, would you suggest an AI model to help you relax?"

[2184] 2. Example of using the Emotion Engine:

[2185] When a user visits a virtual store, the emotion engine collects and analyzes the user's emotional data using a webcam and microphone. Based on the analysis results, the display priority of the model is controlled.

[2186] This invention makes it possible to provide an optimal generative AI model that corresponds to the user's emotions, thereby improving the user experience and enabling efficient model utilization.

[2187] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2188] Step 1: User Registration

[2189] Subject: User, Device, Server

[2190] Input: The user enters required information such as email address, password, and username on the account creation page.

[2191] Specific operation: The terminal sends the information entered by the user to the server as an HTTP request. The server validates the received information and, if there are no problems, saves the new user information in the database.

[2192] Output: The server returns a successful registration message to the user.

[2193] Step 2: User authentication

[2194] Subject: User, Device, Server

[2195] Input: The user enters their email address and password on the login page.

[2196] Specific operation: The terminal sends the entered authentication information to the server. The server searches the database for user information and performs authentication if there is a match.

[2197] Output: The server grants the user access to the market with a message of successful authentication.

[2198] Step 3: Model exploration

[2199] Subject: User, Device, Server

[2200] Input: A user enters a search query to explore generative AI models in a specific area within the market.

[2201] How it works: The device sends a search query to the server, which retrieves a list of relevant generative AI models from the database and analyzes the user's emotions using the emotion engine.

[2202] Output: The server returns a list of generated AI models that correspond to the user. Based on the emotion engine, models that match the user's emotions are displayed preferentially.

[2203] Step 4: Model selection

[2204] Subject: User, Device, Server

[2205] Input: The user selects the generative AI model of interest.

[2206] Specific operation: The device requests detailed information about the selected model. The server retrieves the details from the database and sends them to the device.

[2207] Output: The device displays detailed information about the generated AI model to the user.

[2208] Step 5: Purchase a model

[2209] Subject: User, Device, Server

[2210] Input: The user clicks the purchase button and enters payment information.

[2211] Specific operation: The device sends a purchase request and payment information to the server. The server processes the payment and, if successful, grants the user access to the generated AI model.

[2212] Output: The server sends the user a message confirming the purchase and providing an API key and download link.

[2213] Step 6: Use the model

[2214] Subject: User, Device, Server

[2215] Input: Uses a generative AI model purchased by the user.

[2216] Specific operation: The user accesses the generated AI model through their device. The server processes the API request and provides the required data and model to the user.

[2217] Output: The user uses the generative AI model for their own purposes.

[2218] Step 7: Customization Request

[2219] Subject: User, Device, Server

[2220] Input: The user details the customization requirements for the generative AI model.

[2221] How it works: The device sends a customization request to the server, which then forwards the request to a specialized data scientist.

[2222] Output: The server notifies the user when a new tuned model is available.

[2223] Step 8: Monitoring and Maintenance

[2224] Subject: Server

[2225] Input: Usage data for generative AI models

[2226] Specific operation: The server monitors the usage of the generative AI model in real time, analyzes the collected data, and if an abnormality is detected, notifies the user and performs necessary maintenance.

[2227] Output: The server evaluates the suitability of the model for use based on the results of the analysis of the usage data and makes corrections if necessary.

[2228] 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.

[2229] 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.

[2230] 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.

[2231] [Fourth embodiment]

[2232] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[2233] 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.

[2234] 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).

[2235] 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.

[2236] 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.

[2237] 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).

[2238] 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.

[2239] 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.

[2240] 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.

[2241] 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.

[2242] 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.

[2243] 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.

[2244] 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."

[2245] A system for providing generative AI models specialized for specific fields is configured as follows.

[2246] System Overview

[2247] This system builds a marketplace that provides generative AI models specialized in specific fields, and allows users to access the marketplace to search, purchase, and use models. The system mainly consists of the following components:

[2248] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[2249] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes databases, authentication systems, and payment systems.

[2250] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, and transaction history.

[2251] Explanation of program processing

[2252] User Registration and Authentication

[2253] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[2254] Terminal: Sends the entered information to the server as an HTTP request.

[2255] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[2256] Server: Returns a successful registration message to the user.

[2257] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[2258] Model Exploration and Selection

[2259] Users: Explore generative AI models optimized for specific domains (e.g., healthcare, finance, etc.).

[2260] Device: Sends the user's search query to the server.

[2261] Server: Retrieves a list of relevant generative AI models from the database and returns them to the user.

[2262] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[2263] Server: Retrieves detailed information from the database and sends it to the device.

[2264] Terminal: Display detailed information to the user.

[2265] Purchasing and using models

[2266] User: Select the model they wish to purchase and enter their payment information to complete the purchase.

[2267] Terminal: Sends the entered payment information to the server.

[2268] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[2269] Users: Download purchased models or access them through API.

[2270] Model tuning and customization

[2271] User: If a purchased model needs customization, the user sends a request to the server with their specific requirements.

[2272] Server: Receives requests and relays them to expert data scientists.

[2273] Server: Data scientists tune the model and generate a new model.

[2274] Server: When a new model is ready, it notifies the user and provides access.

[2275] Monitoring and Maintenance

[2276] Server: Monitors the usage of the generative AI model in real time.

[2277] Server: Collects and analyzes usage data and notifies users if anomalies are detected, such as when a model responds excessively slowly or when the error rate is high.

[2278] Server: Once an abnormality is confirmed, maintenance work will be carried out and users will be notified that the service is available again once the problem has been fixed.

[2279] Specific examples

[2280] 1. User Registration and Authentication Example

[2281] Researchers at medical institutions register an account to use the system for the first time.

[2282] Researchers log in and search for generative AI models to find AI models specialized for pathological diagnosis.

[2283] 2. Example of model purchase

[2284] Researchers check detailed information about the pathology diagnostic AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[2285] After purchase, researchers can download the model and begin using it in their actual research.

[2286] 3. Model Tuning Example

[2287] A researcher requests customization, requesting detailed tuning based on clinical data.

[2288] Specialized data scientists will handle the process and provide newly tuned models.

[2289] This invention makes it possible to safely and effectively provide generative AI models specialized for specific fields, enabling users to expect high effectiveness in those fields. In addition, the ease of customization and maintenance improves user convenience.

[2290] The processing flow will be explained below.

[2291] User Registration and Authentication

[2292] Step 1:

[2293] User: Visit the Market and open the account creation page.

[2294] Step 2:

[2295] Device: Enter the required information (email address, password, username, etc.).

[2296] Step 3:

[2297] Terminal: Sends the entered information to the server as an HTTP request.

[2298] Step 4:

[2299] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[2300] Step 5:

[2301] Server: Returns a successful registration message to the user.

[2302] Step 6:

[2303] User: Opens the login page and enters their email address and password.

[2304] Step 7:

[2305] Terminal: Sends the entered authentication information to the server.

[2306] Step 8:

[2307] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[2308] Step 9:

[2309] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[2310] Model Exploration and Selection

[2311] Step 1:

[2312] User: Select a specific sector within the market (e.g., healthcare, finance, etc.).

[2313] Step 2:

[2314] On your device: Enter your search query in the search bar within Market.

[2315] Step 3:

[2316] Device: Sends a search query to the server.

[2317] Step 4:

[2318] Server: Obtain a list of relevant generative AI models from the database.

[2319] Step 5:

[2320] Server: Returns the retrieved model list to the user.

[2321] Step 6:

[2322] User: Select the model of interest from the returned list of models.

[2323] Step 7:

[2324] User: Sends a request to view detailed information about a selected model.

[2325] Step 8:

[2326] Server: Retrieves detailed information from the database and sends it to the device.

[2327] Step 9:

[2328] Terminal: Display detailed information to the user.

[2329] Purchasing and using models

[2330] Step 1:

[2331] User: Clicks the purchase button and enters payment information.

[2332] Step 2:

[2333] Terminal: Sends purchase request and payment information to the server.

[2334] Step 3:

[2335] Server: Processes the payment, and if successful, sends the user a purchase confirmation message and grants them access to the model.

[2336] Step 4:

[2337] Users: Download purchased models or access them via API.

[2338] Model tuning and customization

[2339] Step 1:

[2340] Users: If you require customization of the purchased model, please describe your request in detail.

[2341] Step 2:

[2342] Device: Sends a customization request to the server.

[2343] Step 3:

[2344] Server: Forwards the request to an expert data scientist.

[2345] Step 4:

[2346] Server: Data scientists perform tuning and create new generative AI models.

[2347] Step 5:

[2348] Server: Saves the new generative AI model in the database and notifies the user.

[2349] Step 6:

[2350] User: Use the improved model with the new access information.

[2351] Monitoring and Maintenance

[2352] Step 1:

[2353] Server: Monitors the usage of the generative AI model in real time.

[2354] Step 2:

[2355] Server: Collects and analyzes usage data.

[2356] Step 3:

[2357] Server: If an anomaly is detected, a notification is sent to the user.

[2358] Step 4:

[2359] Users: Receive notifications and take action as needed.

[2360] Step 5:

[2361] Server: If a problem is identified, maintenance work is carried out to fix it.

[2362] Step 6:

[2363] Server: Once the fix is ​​complete, notify users that it is available again.

[2364] In this way, the system provides generative AI models specialized for specific fields and can be effectively operated to meet the diverse needs of users.

[2365] Example 1

[2366] 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."

[2367] Platforms that provide generative AI models specialized in specific fields are required to build an environment where users can efficiently search, purchase, and use generative AI models. However, many current systems lack sufficient functionality for user authentication, search query analysis, customization request response, usage monitoring, and anomaly detection, making it difficult to improve user convenience and reliability.

[2368] 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.

[2369] In this invention, the server includes means for building a platform that provides generative AI models specialized in specific fields, means for storing generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for generative AI models, means for adjusting generative AI models in response to customization requests, means for monitoring the usage status of generative AI models and performing necessary maintenance work, means for users to log in and input authentication information, and means for analyzing user search queries and suggesting appropriate generative AI models. This enables users to efficiently search for, purchase, customize, and use generative AI models.

[2370] "Specific fields" refer to areas with specific expertise or needs, such as medicine, finance, or education.

[2371] "Generative AI models" refer to artificial intelligence models that generate new data and content using techniques such as generative adversarial networks (GANs) and natural language generation (NLG).

[2372] "Platform" refers to the infrastructure that provides an online environment for users to search for, purchase, and use generative AI models.

[2373] "Database" refers to a system for systematically storing and managing generative AI models, related user information, transaction history, etc.

[2374] "End User" refers to the ultimate user who utilizes the generative AI model to provide a specific application or service.

[2375] A "search query" refers to a keyword or phrase entered by a user when searching for a generative AI model.

[2376] "Customization Request" means a request submitted by a User to adjust or refine a Generative AI Model based on their specific needs.

[2377] "Maintenance work" refers to regular checks and maintenance work to ensure that generative AI models function properly.

[2378] "Login" refers to the authentication procedure required for a user to access a system.

[2379] "Authentication Information" refers to information such as email address and password that a User provides to prove access to a System.

[2380] "Search query analysis" refers to the process of analyzing a search query entered by a user and proposing the optimal generative AI model.

[2381] The system of the present invention builds a platform that provides generative AI models specialized in specific fields, and allows end users to search, purchase, customize, and use generative AI models using this platform. The main components of the system are as follows:

[2382] 1. User Device

[2383] A device that users access and operate. It can be a PC, smartphone, tablet, etc. Users access the system through their device to register an account, log in, search for models, view detailed information, make purchases, and request customization.

[2384] 2. Server

[2385] The server is the central component that manages and operates the entire system. It stores generative AI models in a database and includes an authentication system for user authentication, a search query analysis system, a payment system, and a system for responding to customization requests. It also provides API endpoints for users to access. The server communicates with the database and returns appropriate information based on the user's request.

[2386] 3. Database

[2387] This is a system for storing data such as generative AI models, user information, and transaction history. This database uses database systems such as PostgreSQL and Elasticsearch, particularly to speed up searches and maintain data consistency.

[2388] For example, a user accesses a market website and enters the required information (email address, password, username, etc.) on the account registration page. The device sends this information to the server, which validates it and stores it in the database. When the user enters their authentication information on the login page, the server authenticates the user and, if successful, allows them to access the market.

[2389] Next, the user searches for generative AI models suitable for a specific field (e.g., medicine or finance). The device sends the search query to the server, which retrieves and returns a list of relevant models from the database. The user selects the model of interest and requests that detailed information be displayed. The server retrieves the detailed information from the database and sends it to the device, allowing the user to purchase or customize the model.

[2390] Furthermore, if a user requests customization, the request is forwarded to the server and handled by a specialized data scientist. Once customization is complete, a new model is generated and provided to the user. The server also monitors the usage of the generated AI model in real time and takes appropriate action if an abnormality is detected.

[2391] Prompt Sentence Examples

[2392] 1. Researchers at medical institutions must register an account to use the system for the first time.

[2393] 2. Purchase and download the AI ​​model to be used for pathology diagnosis.

[2394] 3. Customize the pathology diagnostic AI model you purchased based on clinical data.

[2395] In this way, the system effectively provides domain-specific generative AI models, enhancing user convenience.

[2396] The flow of the identification process in the first embodiment will be described with reference to FIG.

[2397] Step 1: User Registration

[2398] Users: Visit the Market registration page and enter your name, email address, and password.

[2399] Input: Name, email address, and password entered by the user into the form.

[2400] Output: Registration successful message.

[2401] Terminal: Sends user input information to the server in JSON format.

[2402] Input: Information entered by the user.

[2403] Output: HTTP POST request in JSON format.

[2404] Server: Validate the received information using a validation library (e.g. Joi) and, if there are no problems, save the user information to the PostgreSQL database.

[2405] Input: User information in JSON format.

[2406] Output: The new user information is saved in the database.

[2407] Server: Returns a successful registration message to the user in JSON format.

[2408] Input: User information saved successfully.

[2409] Output: Registration successful message.

[2410] Step 2: User Login

[2411] User: Access the login page and enter the registered email address and password.

[2412] Enter your email address and password.

[2413] Output: Login request.

[2414] Terminal: Sends the user's authentication information to the server in a JSON-formatted HTTP POST request.

[2415] Input: The credentials entered by the user.

[2416] Output: HTTP POST request in JSON format.

[2417] Server: Validates the received authentication information using a password authentication system (e.g. bcrypt) and issues a JWT token if it matches.

[2418] Input: User credentials.

[2419] Output: JWT token.

[2420] Server: Sends the JWT token to the user terminal.

[2421] Input: JWT token.

[2422] Output: User authentication success message and JWT token.

[2423] Step 3: Model Search

[2424] Users: Find generative AI models optimized for specific domains through search queries.

[2425] Input: Search query.

[2426] Output: The search request.

[2427] Terminal: Sends the user's query to the server as an HTTP GET request in JSON format.

[2428] Input: Search query.

[2429] Output: HTTP GET request in JSON format.

[2430] Server: Uses a search engine such as Elasticsearch to retrieve the relevant generative AI model from the database.

[2431] Input: Search query.

[2432] Output: A list of generative AI models.

[2433] Server: Returns search results to the user's device.

[2434] Input: A list of generative AI models.

[2435] Output: Search results.

[2436] Step 4: View model details

[2437] User: Select the model they are interested in and request more information.

[2438] Input: Model ID.

[2439] Output: Model details request.

[2440] Device: Send an HTTP GET request in JSON format containing the model ID to the server.

[2441] Input: Model ID.

[2442] Output: HTTP GET request in JSON format.

[2443] Server: Obtains detailed information about the selected generative AI model from the database and sends it to the user's device.

[2444] Input: Model ID.

[2445] Output: Detailed information about the model.

[2446] Terminal: Displays the received detailed information to the user.

[2447] Input: Model details.

[2448] Output: The detailed information displayed to the user.

[2449] Step 5: Purchase a model

[2450] User: Selects the model they want to purchase and enters their payment information to begin the checkout process.

[2451] Input: Payment information.

[2452] Output: Purchase request.

[2453] Terminal: Sends the user's payment information to the server via an HTTPS POST request in JSON format.

[2454] Input: Payment information.

[2455] Output: HTTPS POST request in JSON format.

[2456] Server: Process the payment using the Stripe API, and if successful, record the purchase confirmation message and model access rights in the database.

[2457] Input: Payment information.

[2458] Output: Purchase successful message and updated access rights.

[2459] Server: Sends a purchase confirmation message and access right information to the user terminal.

[2460] Input: Purchase success message and access rights information.

[2461] Output: Notification to user terminal.

[2462] Step 6: Use the model

[2463] Users: Access or download purchased models via API.

[2464] Input: Access request.

[2465] Output: Model use.

[2466] Server: Receives requests to access a model, performs appropriate authentication, and returns the model file or issues an API key.

[2467] Input: Access request.

[2468] Output: Model file or API key.

[2469] Terminal: Provide the received model file or API key to the user.

[2470] Input: Model file or API key.

[2471] Output: The model is made available to the user.

[2472] (Application example 1)

[2473] 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."

[2474] Modern factories require efficient and rapid optimization of manufacturing processes. It is difficult to easily obtain and use generative AI models suited to specific manufacturing environments. Furthermore, customization of generative AI models and effective monitoring after their deployment are required. To address these challenges, an effective means is needed to easily provide AI models for optimizing the operation of factory robots, and to customize and maintain those models.

[2475] 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.

[2476] In this invention, the server includes means for building a marketplace that provides generative AI models specialized in specific fields, means for storing the generative AI models in a database, means for end users to access and search for models and display detailed information, means for purchasing and paying for the generative AI models, means for tuning the generative AI models in response to customization requests, means for monitoring the usage status of the generative AI models and performing necessary maintenance, means for end users to select, download, or integrate generative AI models to be applied to factory robots, and means for optimizing the operation of the factory robots for specific manufacturing processes. This enables end users to easily obtain generative AI models suitable for their manufacturing environment and effectively apply and operate them.

[2477] A "generative AI model" is an algorithmic model of artificial intelligence that is generated for a specific field or application, and exhibits performance optimized for that field.

[2478] "Market" means the online platform where end users can find, purchase, customize, download, and integrate Generative AI Models.

[2479] A "database" is an information recording system for managing and storing generative AI models, user information, transaction history, etc.

[2480] "End users" are the final users of generative AI models, such as factory workers and managers.

[2481] "Customization Request" means a request by an End User to tailor a Generative AI Model based on their specific needs and requirements.

[2482] "Tuning" refers to the process of adjusting a generative AI model to optimize it for the end user's specific use and environment.

[2483] "Monitoring" refers to the act of monitoring the usage and performance of generative AI models in real time.

[2484] "Maintenance" refers to maintenance work to maintain the performance and reliability of the generative AI model, including fixing bugs and making improvements.

[2485] A "factory robot" is a mechanical device used to automate specific tasks in manufacturing sites, and its operations are controlled by a program.

[2486] "Manufacturing process" refers to the series of operations or steps that transform raw materials into a finished product, including welding, assembly, quality inspection, etc.

[2487] "Integration" refers to the act of end users applying purchased generative AI models to their own systems or machinery (e.g., factory robots) to make them work together.

[2488] This invention relates to a system that enables end users to easily search, purchase, customize, download, and integrate generative AI models suited to specific manufacturing processes. This system is primarily composed of the following hardware and software:

[2489] Hardware and Software Use Cases

[2490] Hardware: A factory robot (e.g., a generic robotic arm) and the smartphone or tablet (iOS or Android device) that controls it.

[2491] software:

[2492] Robot control software (e.g., Robot Operating System (ROS)).

[2493] A cloud service (e.g., AWS or Google Cloud) as a model management server.

[2494] The application for operation on smart devices will be developed using React Native.

[2495] Program processing explanation

[2496] In this invention, the server, terminal, and user work together to search, select, purchase, customize, and use generative AI models. The specific processing steps are as follows:

[2497] 1. User Registration and Authentication

[2498] Users open the app on their smartphone and enter the required information on the account creation page.

[2499] The terminal transmits the input information to the server as an HTTP request.

[2500] The server validates the received information and, if there are no problems, saves the new user information in the database and returns a message to the user confirming successful registration.

[2501] 2. Model Search and Selection

[2502] Users enter a search query to explore generative AI models specific to their manufacturing process.

[2503] The terminal transmits the user's search query to the server.

[2504] The server retrieves a list of relevant generative AI models from the database and returns it to the user.

[2505] The user selects a model of interest from the returned list and requests that its detailed information be displayed.

[2506] The server retrieves the detailed information from the database and sends it to the terminal.

[2507] The terminal displays detailed information to the user.

[2508] 3. Purchasing and Using the Model

[2509] Users select the generative AI model they want to purchase and enter their payment information.

[2510] The terminal transmits the entered payment information to the server.

[2511] The server processes the payment and, if successful, gives the user a purchase confirmation message and access to the model.

[2512] Users can download the purchased model or integrate it into their robot control system via API.

[2513] 4. Tuning and customizing the model

[2514] When a user desires tuning specialized for a specific production line, the user writes a customization request and sends it to the server.

[2515] The server receives the request and relays it to the data scientist.

[2516] The server will notify the user and grant access as soon as a new tuned model is available.

[2517] 5. Monitoring and Maintenance

[2518] The server monitors the usage of the generative AI model in real time.

[2519] The server collects and analyzes usage data and notifies the user if an anomaly is detected, such as a robot slowing down or an increased error rate.

[2520] The server will perform maintenance when an abnormality is detected and notify users that the service is available again once the problem has been fixed.

[2521] Specific examples

[2522] For example, if an end user wants to apply an AI model to a welding process in a factory, they might enter a prompt like this:

[2523] "Optimize the welding seam angle and speed used by this robotic arm."

[2524] This allows the generative AI model to optimize the parameters of the welding process, reducing error rates and improving production speed.

[2525] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[2526] Step 1: User Registration and Authentication

[2527] User: Opens the smartphone app and enters the required information (email address, password, factory ID, etc.) on the account creation page.

[2528] Terminal: Sends the entered information to the server as an HTTP request.

[2529] Server: Validates the received information and, if there are no problems, saves the new user information to the database. If the save is successful, generates a message indicating successful registration and returns it to the terminal. Examples of validation include detecting invalid email addresses and passwords that are too short.

[2530] Step 2: Log in

[2531] User: Enters email address and password on the login page and clicks the login button.

[2532] Terminal: Sends the entered information to the server as an HTTP request.

[2533] Server: Searches for the corresponding user information in the database and compares it with the authentication information. If it matches, it generates a login success message and session information and sends them to the terminal.

[2534] Step 3: Model exploration

[2535] User: Enter a specific manufacturing process (e.g., welding, assembly) into the app's search bar and press the search button.

[2536] Terminal: Sends the search query as an HTTP request to the server.

[2537] Server: Searches the database for a list of relevant generative AI models, generates a list of relevant models, and sends it to the device. Specifically, it filters the models by tags and keywords related to the manufacturing process.

[2538] Step 4: Model selection and detailed display

[2539] User: Selects the generative AI model of interest from the returned list and requests more information about it.

[2540] Device: Send the model ID to the server as an HTTP request.

[2541] Server: Retrieves detailed information about the selected model from the database, generates detailed information, and sends it to the terminal. The detailed information includes the model's scope of application, expected results, and usage methods.

[2542] Step 5: Purchase and pay for the model

[2543] User: Select the generative AI model they want to purchase and enter their payment information (credit card number, security code, etc.).

[2544] Terminal: Sends payment information to the server as an HTTP request.

[2545] Server: Works with the payment processing system to perform payment, and if successful, generates a purchase confirmation message and access rights for the model and sends them to the terminal. If unsuccessful, generates an error message and sends it to the terminal.

[2546] Step 6: Download and integrate the model

[2547] User: Presses a button to download the purchased generative AI model.

[2548] Device: Sends a model download request to the server.

[2549] Server: Retrieves the corresponding model file from the database, generates a download link, and sends it to the device.

[2550] Users: Click the download link and integrate the model file into their robot control system. Specifically, they load the model using software such as ROS and set parameters to optimize the robot's behavior.

[2551] Step 7: Tune and customize the model

[2552] User: If a user wants specialized tuning for a specific production line, he or she enters a customization request and sends it to the server.

[2553] Terminal: Sends the customization request to the server as an HTTP request.

[2554] Server: Receives requests and forwards them to data scientists, who tune the model based on the requests and generate new versions of the model.

[2555] Server: Saves the new tuned model to the database and notifies the user of the update.

[2556] Step 8: Monitoring and Maintenance

[2557] Server: Collects logging and performance metrics to monitor the usage of generative AI models in real time.

[2558] Server: If abnormal behavior (e.g., delayed response, increased error rate) is detected, the server identifies the anomaly based on an anomaly detection algorithm and notifies the user. Statistical anomaly detection techniques and machine learning algorithms are used for anomaly detection.

[2559] Server: If an anomaly is detected, the maintenance team is notified and the model is available again once the fix is ​​complete. Continuous monitoring ensures the reliability and effectiveness of the model.

[2560] 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.

[2561] The system for providing generative AI models specialized for specific fields is configured as follows: This system also incorporates an emotion engine that recognizes user emotions.

[2562] System Overview

[2563] This system creates a marketplace that provides generative AI models specialized in specific fields, allowing users to access the marketplace and search, purchase, and use models. It also has the ability to recognize user emotions and suggest models and interact with them based on those emotions.

[2564] The system mainly consists of the following components:

[2565] 1. User terminal: A device that a user accesses and operates. This includes PCs, smartphones, tablets, etc.

[2566] 2. Server: This is the central component that manages and stores generative AI models and interacts with user devices to execute operations. It also includes a database, authentication system, payment system, emotion engine, etc.

[2567] 3. Database: A system that stores data such as generative AI models specialized in specific fields, user information, transaction history, and emotional data.

[2568] 4. Emotion Engine: A component for recognizing and analyzing user emotions.

[2569] Explanation of program processing

[2570] User Registration and Authentication

[2571] User: Access the market and enter the required information (email address, password, username, etc.) on the account creation page.

[2572] Terminal: Sends the entered information to the server as an HTTP request.

[2573] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[2574] Server: Returns a successful registration message to the user.

[2575] The user then enters their credentials on the login page, and the server authenticates the user. If authentication is successful, the user is allowed to access the market.

[2576] Model Exploration and Selection

[2577] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[2578] Device: Sends the user's search query to the server.

[2579] Server: Retrieves a list of applicable generative AI models from the database and returns it to the user. The emotion engine recognizes the user's emotions and prioritizes models that evoke positive emotional responses.

[2580] User: Selects the model of interest from the returned list of models and requests that its details be displayed on the device.

[2581] Server: Retrieves detailed information from the database and sends it to the device.

[2582] Terminal: Display detailed information to the user.

[2583] Purchasing and using models

[2584] User: Clicks the purchase button and enters payment information.

[2585] Terminal: Sends purchase request and payment information to the server.

[2586] Server: Processes the payment and, if successful, sends the user a purchase confirmation message and grants them access to the model.

[2587] Users: Download purchased models or access them via API.

[2588] Model tuning and customization

[2589] Users: If you require customization of the purchased model, please describe your request in detail.

[2590] Device: Sends a customization request to the server.

[2591] Server: Forwards the request to an expert data scientist.

[2592] Server: Data scientists perform tuning and create new generative AI models.

[2593] Server: Saves the new generative AI model in the database and notifies the user.

[2594] User: Use the improved model with the new access information.

[2595] Monitoring and Maintenance

[2596] Server: Monitors the usage of the generative AI model in real time.

[2597] Server: Collects and analyzes usage data.

[2598] Server: If an anomaly is detected, a notification is sent to the user.

[2599] Users: Receive notifications and take action as needed.

[2600] Server: If a problem is identified, maintenance work is carried out to fix it.

[2601] Server: Once the fix is ​​complete, notify users that it is available again.

[2602] Incorporating an emotion engine

[2603] Server: Builds an emotion engine and collects emotion data in real time from users' voices, text, facial expressions, etc.

[2604] Server: The emotion engine analyzes the collected data and determines the emotional state.

[2605] Server: Based on the acquired emotion data, the server presents a list of generative AI models that are optimal for the user. For example, if the user is feeling stressed, the server prioritizes models that are effective in reducing stress.

[2606] Server: The emotion engine periodically analyzes the emotion data and suggests or customizes models as needed.

[2607] User: Review the sentiment-based recommendation model and make a purchase or use it.

[2608] Specific examples

[2609] 1. User Registration and Authentication Example

[2610] A pharmaceutical researcher from a healthcare company registers an account to use the system for the first time.

[2611] Researchers log in and search generative AI models to find AI models specialized for specific diseases.

[2612] 2. Example of model purchase

[2613] Researchers check detailed information about the disease-specific AI model, determine that its performance is suitable for their institution, and then decide to purchase it.

[2614] After purchase, researchers can download the model and begin using it in their research.

[2615] 3. Model Tuning Example

[2616] A researcher requests customization, requesting detailed tuning based on clinical data.

[2617] Specialized data scientists will handle the process and provide newly tuned models.

[2618] 4. Example of an Emotion Engine

[2619] As researchers use the system, the emotion engine analyzes their context and recognizes when they are feeling stressed.

[2620] Based on the emotion engine, we present effective guidelines for stress reduction and propose a model to improve researchers' work efficiency.

[2621] This invention enables the safe and effective provision of generative AI models specialized for specific fields. In addition, by incorporating an emotion engine, it is possible to provide more personalized services according to the user's emotional state.

[2622] The processing flow will be explained below.

[2623] User Registration and Authentication

[2624] Step 1:

[2625] User: Visit the Market and open the account creation page.

[2626] Step 2:

[2627] Device: Enter the required information (email address, password, username, etc.).

[2628] Step 3:

[2629] Terminal: Sends the entered information to the server as an HTTP request.

[2630] Step 4:

[2631] Server: Validates the received information and, if there are no problems, saves the new user information to the database.

[2632] Step 5:

[2633] Server: Returns a successful registration message to the user.

[2634] Step 6:

[2635] User: Opens the login page and enters their email address and password.

[2636] Step 7:

[2637] Terminal: Sends the entered authentication information to the server.

[2638] Step 8:

[2639] Server: Searches the database for the relevant user information and checks whether it matches the entered password.

[2640] Step 9:

[2641] Server: If the verification is successful, generate a session ID and return it to the user. If unsuccessful, return an error message.

[2642] Model Exploration and Selection

[2643] Step 1:

[2644] Users: Explore generative AI models within the market that are optimized for specific domains (e.g., healthcare, finance, etc.).

[2645] Step 2:

[2646] On your device: Enter your search query in the search bar within Market.

[2647] Step 3:

[2648] Device: Sends a search query to the server.

[2649] Step 4:

[2650] Server: Obtain a list of relevant generative AI models from the database.

[2651] Step 5:

[2652] Server: Returns the acquired model list to the user device.

[2653] Step 6:

[2654] User: Select the model of interest from the returned list of models.

[2655] Step 7:

[2656] User: Sends a request to view detailed information about a selected model.

[2657] Step 8:

[2658] Server: Retrieves detailed information from the database and sends it to the device.

[2659] Step 9:

[2660] Terminal: Display detailed infor...

Claims

1. A means to build a marketplace that offers specialized generative AI models for specific fields, and A means of storing the generative AI model in a database; and A means for end users to access, search for, and view detailed information about models; A means of purchasing and paying for generative AI models; A means for tuning the generative AI model according to customization requests; and A means to monitor the usage of generative AI models and perform necessary maintenance; A system including:

2. A means to provide detailed information about domain-specific generative AI models; A means for forwarding customization requests from users to experts; and further comprising means for re-serving the tuned generative AI model. The system of claim 1 .

3. A means to analyze the usage of the generated AI model and notify the user if an abnormality is detected, Further, include a means to evaluate the suitability and effectiveness of the model for use; The system of claim 1 .

Images