System

Abstract

A system is provided.SOLUTION: This system is provided with a means for introducing general purpose artificial intelligence and learning expert knowledge and skill for coping with a specific job, a means for evolving the general purpose artificial intelligence to artificial intelligence corresponding to the specific job based on knowledge and guidance provided by an expert in the field, and a means for using an expert book, an article, a record and an expert glossary as the learning data of the general purpose artificial intelligence.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] Currently, general-purpose AI has human-like general knowledge but lacks the specialized capabilities required to perform specific tasks. Therefore, there is a need to develop AI with specialized knowledge and skills for specific industries and tasks. Furthermore, current technology makes this training process inefficient and lacks a way to effectively incorporate the knowledge of industry experts. The present invention aims to solve these problems and provide a system for efficiently training general-purpose AI into AI capable of specialized tasks. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for introducing a general-purpose AI and learning specialized knowledge and skills for a specific job, a means for evolving the general-purpose AI into an AI suitable for the specific job based on knowledge and guidance provided by industry experts, and a means for using specialized books, papers, records, and glossaries as learning data for the general-purpose AI. Specifically, the system further includes a means for selecting a specialized AI model for the specific job and training the model, a means for saving the trained model and generating identification information for the model, and a means for providing the identification information to a user. The system also includes a means for collecting data entered by a user and sending the data to a server, a means for the server to analyze the received data and start training the AI ​​based on the data, and a means for returning the training results to the user.

[0006] "General artificial intelligence" is artificial intelligence that has human-like general knowledge and basic cognitive abilities and can perform a wide range of tasks, not just for specific purposes.

[0007] "Specialized knowledge" is a general term for the detailed and advanced information and skills required in a particular job or industry.

[0008] "Skills" are the specialized abilities and techniques required to perform a particular job or activity.

[0009] "Training data" refers to a data set used in the training process of artificial intelligence, from which the AI ​​learns knowledge and skills.

[0010] A "specialized book" is a book that contains specialized content related to a specific field or job.

[0011] A "paper" is a piece of writing that summarizes the research results and opinions of researchers and experts on a specific topic.

[0012] A "record" is a document or data that describes a particular fact or event.

[0013] A "terminology glossary" is a list or book that collects technical terms and their definitions that are not commonly used in a particular field.

[0014] An "expert" is a person who has advanced knowledge and skills in a particular field or occupation and is considered an authority.

[0015] "Training" is the process by which artificial intelligence acquires knowledge and skills based on given learning data and improves its capabilities.

[0016] A "model" is a mathematical or logical structure that is built based on the learning of artificial intelligence and serves as the basis for performing a specific task.

[0017] "Storage" refers to the act of recording trained artificial intelligence models and data on a storage device or server and keeping them available for future use.

[0018] "Identification information" refers to unique information or a code that identifies a saved model or data.

[0019] "User" means any person or organization that uses the Platform or System to train or otherwise operate AI models.

[0020] A "server" is a computer system that provides data and services over a network.

[0021] "Data preprocessing" refers to a series of operations or techniques that convert training data into a format that is easy for an artificial intelligence model to understand. [Brief explanation of the drawings]

[0022] [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

[0023] 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.

[0024] First, the terms used in the following description will be explained.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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."

[0030] [First embodiment]

[0031] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0032] 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.

[0033] 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).

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0039] 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.

[0040] 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.

[0041] 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.

[0042] 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."

[0043] This invention relates to a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The operation of the system is explained below.

[0044] User operations

[0045] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a translation AI in the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, users can click the "Train AI" button to start the AI ​​training request.

[0046] Device behavior

[0047] The terminal collects the data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," and "Model Type." The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0048] Server Operation

[0049] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," and "Model Type." The server selects a model for training and starts training the AI ​​model based on the received data.

[0050] Training an AI model

[0051] The server selects a neural network model for training based on the received "Model Type." For example, for a translation model, it configures and applies a neural network with specific input and output sizes. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0052] Training results storage and notification

[0053] Once training is complete, the server stores the trained model along with a unique identifier that will be used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[0054] Displaying results on a terminal

[0055] When the device receives the response from the server, it analyzes the response and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message such as "Training successful. Model ID: model12345" may be displayed.

[0056] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[0057] The processing flow will be explained below.

[0058] Step 1:

[0059] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[0060] Step 2:

[0061] The user clicks the "Train AI" button.

[0062] Step 3:

[0063] The device calls the JavaScript trainAI function.

[0064] Step 4:

[0065] The terminal collects the input data of "Industry" and "Model Type".

[0066] Step 5:

[0067] The data collected by the device is converted into JSON format. For example, it is converted into the following format.

[0068] json

[0069] {

[0070] "industry": "Medical",

[0071] "domain_data": [], / / Include medical related data here

[0072] "model_type": "translator"

[0073] }

[0074] Step 6:

[0075] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[0076] Step 7:

[0077] The server receives the POST request and parses the JSON data.

[0078] Step 8:

[0079] The server extracts the industry, domain_data, and model_type fields from the parsed data.

[0080] Step 9:

[0081] The server calls the train_model function to start training the AI ​​model.

[0082] Step 10:

[0083] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[0084] Step 11:

[0085] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[0086] Step 12:

[0087] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[0088] Step 13:

[0089] After the server completes training, it saves the model to storage.

[0090] Step 14:

[0091] The server assigns a unique identification (model ID) to the saved model.

[0092] Step 15:

[0093] The server generates a JSON response containing a training success message and the model ID.

[0094] json

[0095] {

[0096] "message": "Training successful",

[0097] "model_id": "model12345"

[0098] }

[0099] Step 16:

[0100] The server generates a JSON response and sends it back to the device.

[0101] Step 17:

[0102] The device receives a JSON response from the server.

[0103] Step 18:

[0104] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[0105] Example 1

[0106] 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."

[0107] An appropriate training process is necessary for general-purpose AI to effectively learn specialized knowledge and skills for specific tasks and evolve into AI with advanced specialized knowledge. However, conventional methods have not been able to achieve efficient training because they do not effectively utilize user input data and do not fully reflect expert knowledge and guidance. Furthermore, training results are not returned to users quickly and reliably, making them less convenient to use in actual tasks.

[0108] 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.

[0109] In this invention, the server includes: a means for a user to select a specific job or industry and initiate a training request; a means for a terminal to collect user input data, format it in JSON format, and send it to the server; a means for the server to analyze the received data, select a specific model, and initiate AI training based on the data; a means for saving the AI ​​model after completion of training and generating unique identification information; and a means for providing the user with a training result success message including the identification information. This allows the general AI to effectively learn the specialized knowledge and skills required for a specific job, and makes it possible to quickly and reliably provide the trained AI model to the user.

[0110] "General-purpose artificial intelligence" is artificial intelligence that is not limited to a specific use and can handle a variety of tasks and operations.

[0111] "Specific work" refers to specific tasks or activities related to a particular job, role, or industry.

[0112] "Expertise" refers to specialized knowledge or information related to a particular job or field.

[0113] "Skills" refers to the techniques and abilities required to perform a particular task efficiently and accurately.

[0114] An "expert" is someone who has advanced knowledge and experience in a particular job or field.

[0115] "Training data" refers to the datasets used to train artificial intelligence, including specialized literature, documents, records, glossaries, etc.

[0116] "User" refers to any person or organization that uses the System to request training of the Artificial Intelligence.

[0117] A "terminal" is a device used by a user to operate the system, and includes a computer, a smartphone, etc.

[0118] The "JSON format" is a lightweight data exchange format for structuring and representing data.

[0119] A "server" refers to a computer system that receives and processes requests from users.

[0120] "Analysis" refers to the process of understanding the data received and extracting the necessary information.

[0121] A "neural network model" is a specific algorithmic model used to train artificial intelligence, which is based on neural networks.

[0122] "Preprocessing" refers to the process for converting training data into a trainable form.

[0123] "Training" refers to the process of teaching an artificial intelligence model the ability to handle a specific task.

[0124] "Identification information" refers to information that uniquely identifies a trained artificial intelligence model.

[0125] A "success message" refers to a message that informs the user that the training is complete.

[0126] The present invention is a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The detailed implementation method of the system is described below.

[0127] First, the user accesses the system's operation screen and enters data to select a specific business or industry. Specifically, the user enters the industry name in the "Industry" field and the model type in the "Model Type" field. For example, if the user wants to train a translation AI for the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to start the training request.

[0128] Next, the terminal collects the data entered by the user and formats it in JSON format. For example, it generates JSON data containing information such as "Industry," "Domain Data," and "Model Type." This data is sent from the terminal to the specified endpoint of the server as a POST request. The terminal automatically formats the data and sends it to the server.

[0129] The server receives the request sent from the terminal and analyzes the included JSON data, which includes "Industry," "Domain Data," and "Model Type." Based on the analysis results, the server selects a model for training and begins training the AI ​​model based on the data. Specifically, the server selects a neural network model according to the "Model Type" and preprocesses the received "Domain Data" to convert it into a trainable format.

[0130] The server then uses the preprocessed data to train the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. Through this process, the AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[0131] Once training is complete, the server stores the trained model along with a unique identifier, which is used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[0132] Finally, the device receives the response from the server, analyzes it, and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message like "Training successful. Model ID: model12345" can be displayed.

[0133] Take the training of medical translation AI as a concrete example. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[0134] Prompt Sentence Examples

[0135] Prompt: I have created a dataset to train a translation AI specialized for the medical field. Now, please train your AI model on this dataset.

[0136] Input data: "Industry: Medical" "Domain Data: Medical books, papers, medical records, terminology" "Model Type: translator"

[0137] Output: The AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[0138] By linking user input with server processing, this system effectively specializes artificial intelligence for specific tasks, allowing users to quickly and efficiently utilize AI models with the necessary expertise and skills.

[0139] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0140] Step 1:

[0141] The user enters data

[0142] Specific behavior:

[0143] The user accesses the system's operation screen and enters data into the fields for selecting the job and industry. Specifically, enter the industry name in the "Industry" field and the model type in the "Model Type" field. For example, enter "Medical" in "Industry" and "Translator" in "Model Type." Then, click the "Train AI" button to start the training request.

[0144] Input: User-entered industry name and model type

[0145] Output: Training request started

[0146] Step 2:

[0147] The device collects and formats the data

[0148] Specific behavior:

[0149] The terminal collects the "Industry," "Domain Data," and "Model Type" data entered by the user. It formats the collected data in JSON format and prepares it for sending to the server. The generated JSON data looks like this, for example:

[0150] {

[0151] "Industry": "Medical",

[0152] "Domain Data": "Medical books, papers, medical records, terminology",

[0153] "Model Type": "translator"

[0154] }

[0155] Input: User-entered industry name, domain data, and model type

[0156] Output: JSON format data

[0157] Step 3:

[0158] The device sends data to the server

[0159] Specific behavior:

[0160] The formatted JSON data is sent as a POST request to the server's / train-ai endpoint. The device checks whether the transmission was successful.

[0161] Input: JSON format data

[0162] Output: Sending a POST request to the server

[0163] Step 4:

[0164] The server analyzes the received data

[0165] Specific behavior:

[0166] The server receives the request sent from the device and parses the included JSON data, extracting the following information from the parsed data: "Industry," "Domain Data," and "Model Type."

[0167] Input: Received JSON data

[0168] Output: Parsed "Industry", "Domain Data", and "Model Type"

[0169] Step 5:

[0170] The server selects the training model and preprocesses the data.

[0171] Specific behavior:

[0172] The server selects an appropriate neural network model based on the extracted "Model Type," and then preprocesses the acquired "Domain Data" and converts it into a trainable format.

[0173] Input: Parsed "Model Type" and "Domain Data"

[0174] Output: Data converted into a trainable format

[0175] Step 6:

[0176] The server trains the AI ​​model

[0177] Specific behavior:

[0178] Using the preprocessed data, the server trains the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. In this process, the AI ​​model acquires skills specific to medical translation.

[0179] Input: Data converted into a trainable format

[0180] Output: A trained AI model

[0181] Step 7:

[0182] The server stores and notifies the training results

[0183] Specific behavior:

[0184] Once training is complete, the server saves the model and generates a unique identifier (Model ID). The server generates a response containing a training success message and the unique identifier and sends it to the device.

[0185] Input: A trained AI model

[0186] Output: Model identification and a success message

[0187] Step 8:

[0188] The device receives the server's response and displays it to the user.

[0189] Specific behavior:

[0190] The device receives the response from the server, analyzes its contents, and displays a message indicating successful training and the model's identification information to the user. For example, a message such as "Training successful. Model ID: model12345" is displayed.

[0191] Input: Response from the server

[0192] Output: Displaying training results to the user

[0193] (Application example 1)

[0194] 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."

[0195] In modern brick-and-mortar stores, it is difficult for customers to quickly obtain detailed product information and usage instructions. Furthermore, there is a lack of a system that properly recommends related products, which prevents customers from making optimal choices when purchasing. Furthermore, inventory cannot be checked in real time, which causes inconvenience for both customers and stores.

[0196] 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.

[0197] In this invention, the server includes: means for introducing a general-purpose artificial intelligence and learning specialized knowledge and skills for a specific job; means for evolving the general-purpose artificial intelligence into an artificial intelligence suitable for the specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence; means for recommending detailed information, usage guides, and related products related to the specific job; and means for reading barcodes to obtain product information. This allows customers to obtain product information, usage instructions, and related product recommendations in real time in stores, improving the accuracy of their purchasing decisions.

[0198] "General-purpose artificial intelligence" is artificial intelligence that is not dependent on specific tasks or jobs and can be applied in a wide range of fields.

[0199] "Specialized knowledge" is detailed and advanced knowledge acquired in a particular job or industry.

[0200] "Technology" refers to the specific techniques and skills required for a particular job or industry.

[0201] An "industry expert" is a professional with extensive experience and in-depth knowledge in a particular industry.

[0202] A "specific job" is a specific job or role performed in a particular industry or field.

[0203] A "specialty book" is a publication that provides advanced knowledge or information in a specific field.

[0204] A "paper" is a document that reports the results of detailed research and analysis on a specific research subject.

[0205] A "record" is information that preserves past facts, events, and data.

[0206] A "terminology glossary" is a dictionary or list of technical terms in a particular field.

[0207] "Detailed information" means specific and comprehensive information about a product or service.

[0208] A "usage guide" is a document that explains the specific steps and methods for using a product or service.

[0209] "Related products" are other products that are highly related to a particular product.

[0210] "Recommendation" is the act of introducing something to others based on specific conditions or requirements.

[0211] A "barcode" is a series of lines and numbers used to represent the identity of a product or item.

[0212] "Product information" is detailed data about the product's characteristics, specifications, usage, price, etc.

[0213] "Stock status" is information that indicates how much of a particular product is currently in stock.

[0214] This invention provides a system that allows customers in physical stores to quickly obtain detailed product information, usage instructions, and related product recommendations. This system employs general artificial intelligence (AI) and is trained to acquire specialized knowledge and skills corresponding to specific tasks.

[0215] Hardware and Software Configuration

[0216] The system's hardware consists of a smartphone, a server, and a barcode reader, while the software uses an API client developed in Python, TensorFlow or PyTorch for training the neural network model, and Flask or Django for providing the REST API server.

[0217] Feature details

[0218] 1. Training the general-purpose AI: The server uses specialized books, papers, records, and terminology to train the general-purpose AI with specialized knowledge and skills for specific tasks. This data is sent to the server in JSON format and used to train the AI ​​model.

[0219] 2. Product information acquisition: Using a smartphone application, the customer scans the product barcode, which retrieves the corresponding product information from the server and displays it on the smartphone screen.

[0220] 3. Providing usage guide: Based on the product ID, the server provides instructions on how to use and assemble the corresponding product. This information is also displayed through the smartphone application.

[0221] 4. Related product recommendation: The server recommends other related products based on the specific product information, and this recommendation information is also displayed on the smartphone screen.

[0222] 5. Check stock availability: The server can check the stock availability of a particular product in real time and provide that information on the smartphone screen.

[0223] Processing flow

[0224] User operation: The user operates the smartphone app and scans the barcode.

[0225] Terminal operation: Collects barcode information and sends it to the server in JSON format.

[0226] Server Action: Parse the product information based on the received barcode information and generate and return a response containing detailed information, usage guides, related product recommendations, and stock availability.

[0227] Displaying the results: The smartphone app receives the response from the server and displays it to the user.

[0228] Specific examples

[0229] For example, if a user wants to get information about the latest running shoes, they can scan the barcode with their smartphone and instantly see detailed product information, instructions for use, recommendations for related running gear, and the product's availability.

[0230] Prompt Sentence Examples

[0231] "I'm looking for the latest running shoes. I want something durable and suitable for long distance running. What are the best options?"

[0232] In this way, an intelligent system that provides advanced support for the shopping experience in physical stores will be realized.

[0233] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0234] Step 1:

[0235] A user launches a smartphone application and scans a barcode. The barcode reader is used to obtain the product's barcode information, which is then entered into the application. This barcode information is a series of numbers and letters that identify a specific product.

[0236] Step 2:

[0237] The device (smartphone) formats the acquired barcode information into JSON format. This data includes the barcode information and the user's device identification information. An HTTP POST request is generated to send the formatted JSON data to the server.

[0238] Step 3:

[0239] The server receives the JSON data sent from the device. The received data includes the product barcode information. The server analyzes this information and retrieves the corresponding product details from the database.

[0240] Step 4:

[0241] The server generates usage guides and related product data based on the acquired product details. Product usage guides are data that includes procedures and operating methods, and related product data is information on other recommended products. These data are processed on the server and converted into JSON format.

[0242] Step 5:

[0243] The server queries the inventory management system to check the real-time availability of a product. The query returns the product's stock quantity and its status. The server also converts this stock information into JSON format.

[0244] Step 6:

[0245] The server compiles the generated product details, usage guide, related product information, and stock status into a single JSON response and sends this response to the device as an HTTP response. The response contains all the necessary product information.

[0246] Step 7:

[0247] The device parses the JSON response received from the server and displays it on the screen. Users can then use the smartphone application to check product details, usage guides, related product recommendations, and stock availability, allowing them to make product selections.

[0248] These steps enable users to enjoy an efficient and high-quality shopping experience in physical stores. For example, when selecting running shoes, users can obtain optimal product information by entering a prompt such as, "I'm looking for the latest running shoes. I want something durable and suitable for long-distance running. Which product would be best?"

[0249] 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.

[0250] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[0251] User operations

[0252] Users first select the specific job or industry they want to give expertise in, entering "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, users click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[0253] Device behavior

[0254] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0255] Server Operation

[0256] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[0257] Training an AI model

[0258] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0259] Training results storage and notification

[0260] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[0261] Displaying results on a terminal

[0262] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[0263] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis certificate, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[0264] The processing flow will be explained below.

[0265] Step 1:

[0266] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[0267] Step 2:

[0268] The user clicks the "Train AI" button.

[0269] Step 3:

[0270] The device calls the JavaScript trainAI function.

[0271] Step 4:

[0272] The terminal collects the input data of "Industry" and "Model Type".

[0273] Step 5:

[0274] The device uses an emotion engine to detect the user's emotional state in real time, for example by analyzing emotions from the user's facial expressions and tone of voice.

[0275] Step 6:

[0276] The device converts the collected data and emotional state into a JSON format, for example, as follows:

[0277] json

[0278] {

[0279] "industry": "Medical",

[0280] "domain_data": [], / / Include medical related data here

[0281] "model_type": "translator",

[0282] "emotional_state": "neutral" / / or "stressed", etc.

[0283] }

[0284] Step 7:

[0285] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[0286] Step 8:

[0287] The server receives the POST request and parses the JSON data.

[0288] Step 9:

[0289] The server extracts the industry, domain_data, model_type, and emotional_state fields from the parsed data.

[0290] Step 10:

[0291] The server calls the train_model function to start training the AI ​​model.

[0292] Step 11:

[0293] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[0294] Step 12:

[0295] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[0296] Step 13:

[0297] The server adjusts the training parameters based on the user's emotional state (emotional_state). For example, if the user is feeling stressed, the difficulty of the training will be reduced.

[0298] Step 14:

[0299] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[0300] Step 15:

[0301] After the server completes training, it saves the model to storage.

[0302] Step 16:

[0303] The server assigns a unique identification (model ID) to the saved model.

[0304] Step 17:

[0305] The server generates a JSON response containing a training success message and the model ID.

[0306] json

[0307] {

[0308] "message": "Training successful",

[0309] "model_id": "model12345"

[0310] }

[0311] Step 18:

[0312] The server generates a JSON response and sends it back to the device.

[0313] Step 19:

[0314] The device receives a JSON response from the server.

[0315] Step 20:

[0316] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[0317] Example 2

[0318] 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."

[0319] Conventional general-purpose AI systems have difficulty effectively learning specialized knowledge and skills specific to specific jobs or industries. Furthermore, because they do not take the user's emotional state into account, the effectiveness of training depends on the user's psychological state, resulting in reduced efficiency. Therefore, a new system that can monitor the user's emotional state in real time and optimize training is needed.

[0320] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: means for learning specialized knowledge and skills for a specific job; means for evolving a general-purpose AI into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as training data; means for monitoring the user's emotional state in real time; means for collecting data entered by the user and formatting the data in JSON format; means for transmitting the formatted data to the server; and means for analyzing the received data and starting AI training based on the data. This enables efficient training of AI specialized for a specific job, and optimal training that takes into account the user's emotional state can be achieved.

[0321] "General-purpose artificial intelligence" is a general-purpose artificial intelligence system that can be used in a wide range of areas, without being limited to specific tasks or applications.

[0322] A "job" refers to a specific task or role that a person performs in their occupation.

[0323] "Skills" is a general term for the techniques and abilities required to perform specific jobs or tasks.

[0324] An "industry expert" refers to an expert with advanced knowledge and experience in a particular industry or field.

[0325] A "specialized book" is a book that provides specialized knowledge related to a particular field or job.

[0326] A "paper" is an academic piece of writing that describes the results of a particular study or investigation.

[0327] "Records" refer to documents or data that record specific facts or events.

[0328] A "terminology glossary" is a collection of technical terms and their definitions used in a particular field or job.

[0329] An "emotion engine" refers to a system or algorithm that monitors and analyzes a user's emotional state in real time.

[0330] "Real-time monitoring" means continuously observing the user's condition in real time and immediately detecting any changes.

[0331] "Data collection" is the act or process of systematically compiling necessary information.

[0332] "JSON format" is an abbreviation for JavaScript Object Notation, and is a lightweight data exchange format for expressing data in text format.

[0333] A "server" is a computer system that receives and responds to requests from clients on a network.

[0334] A "neural network model" is a type of artificial intelligence with a structure that mimics biological nerve cells, and is an algorithm that specializes in learning and prediction for specific tasks.

[0335] "Training" is the process of providing data to an artificial intelligence or machine learning model to learn and improve its predictive capabilities and performance.

[0336] "Training parameters" refer to settings or values ​​that are adjusted during the learning process of an artificial intelligence model.

[0337] "Preprocessing" refers to the preliminary processing of raw data to convert it into a format that is easy for machine learning models to handle.

[0338] "Identification information" refers to information or an ID that uniquely identifies a specific object or data.

[0339] A "prompt" is a document or instruction used as input to a generative artificial intelligence model.

[0340] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills for a specific job, and also combines it with an emotion engine that recognizes the user's emotions. The system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[0341] Users first select the specific job or industry they want to provide expertise in and enter "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, they click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine, which analyzes the user's emotional state and dynamically adjusts training parameters.

[0342] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0343] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[0344] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0345] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[0346] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[0347] Take the training of medical translation AI as a concrete example. When users train a medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[0348] Example prompt sentence:

[0349] "Train a medical translator AI model using data from medical books, research papers, medical records, and medical glossaries. Adjust training parameters based on user's emotional state, ensuring lower difficulty when the user feels stressed."

[0350] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0351] Step 1: User Input

[0352] Users use the system's interface to select the specific job or industry they want to train in by entering "Medical" in the "Industry" field and "translator" in the "Model Type" field. The emotion engine monitors the user's emotional state in real time and collects that data. The input data consists of "Industry," "Model Type," and "Emotional State." For example, if the user is stressed, the "Emotional State" is recorded as "stressed."

[0353] Step 2: Data formatting and transmission by the terminal

[0354] The device collects the "Industry" and "Model Type" input by the user, as well as the "Emotional State" information obtained from the emotion engine, and formats it in JSON format. For example, the following JSON data is generated:

[0355] json

[0356] {

[0357] "Industry": "Medical",

[0358] "Model Type": "translator",

[0359] "Emotional State": "stressed"

[0360] }

[0361] This JSON data is sent as a POST request to the server's / train-ai endpoint. Data formatting and transmission are automated by the device.

[0362] Step 3: Data reception and analysis by the server

[0363] The server receives the POST request sent from the device and parses the included JSON data. The parsed results include "Industry," "Model Type," and "Emotional State." The server first parses the data and obtains the values ​​of each field. For example, "Industry" is "Medical," "Model Type" is "translator," and "Emotional State" is "stressed."

[0364] Step 4: Selecting an AI model and adjusting its parameters

[0365] The server selects an appropriate neural network model for training based on the analyzed data. In this case, because the "Model Type" is "translator," a model specialized for translation is selected. The server also references the "Emotional State" obtained from the emotion engine and adjusts parameters, such as lowering the difficulty of training if the user is feeling stressed.

[0366] Step 5: Preprocess the data and run training

[0367] The server preprocesses the domain data and converts it into a trainable format. For example, it cleans and formats data from medical papers and diagnostic reports. This preprocessed data is then used to train the AI ​​model. Specifically, the data is fed into a neural network, which learns it epoch by epoch.

[0368] Step 6: Save and notify training results

[0369] Once training is complete, the server saves the trained model with a unique identifier (model ID). For example, the trained model is assigned an ID of "model12345". The server then generates a JSON response containing a training success message and the model's identifier, and sends it to the device. An example response is as follows:

[0370] json

[0371] {

[0372] "message": "Training successful",

[0373] "modelID": "model12345"

[0374] }

[0375] Step 7: Displaying the results on your device

[0376] The device receives the response from the server and analyzes its contents. The analysis results include a success message and a model ID. The device displays this information to the user. For example, it might display "Training successful. Model ID: model12345." By checking this result, the user can confirm that the training was successful and obtain the model's identification information.

[0377] (Application example 2)

[0378] 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."

[0379] Conventional general-purpose AI training systems have had the problem of being difficult to efficiently teach specialized knowledge and skills specialized for specific jobs. Furthermore, the lack of feedback based on the user's emotional state can sometimes prevent the training from being fully effective. Furthermore, there was no system that could recommend appropriate content taking the user's emotional state into account, resulting in a less than satisfactory user experience.

[0380] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0381] In this invention, the server includes: a means for introducing a general-purpose artificial intelligence (AI) to learn specialized knowledge and skills for a specific job; a means for evolving the AI ​​into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; a means for using specialized books, papers, records, and technical glossaries as training data for the AI; a means for incorporating an emotion engine that analyzes a user's emotional state in real time and provides appropriate feedback according to that state; and a means for generating prompts and adjusting the training of the AI ​​model based on the emotional state. This enables the AI ​​to efficiently learn specialized knowledge and skills specific to a specific job, and provides feedback according to the emotional state to improve the effectiveness of training. Furthermore, the user experience can be improved by recommending content based on the user's emotional state.

[0382] "General artificial intelligence" is artificial intelligence that can handle a wide range of jobs and tasks.

[0383] "Expertise" refers to detailed information or knowledge about a particular job or field.

[0384] "Skills" refer to the techniques and abilities required to perform a particular job.

[0385] An "industry expert" is a person or group of people who have a high level of expertise in a particular job function or field.

[0386] An "emotion engine" is a software or hardware system for detecting and analyzing a user's emotional state.

[0387] "Real-time" refers to a timeframe in which processing and feedback occurs almost immediately.

[0388] A "prompt" is input data used to give instructions or ask questions to artificial intelligence.

[0389] "Training parameters" are settings or values ​​that can be adjusted during the learning process of an artificial intelligence model.

[0390] A "specialized book" is a book that contains detailed information about a specific job or field.

[0391] A "paper" is an academic document that summarizes research results and considerations on a specific topic.

[0392] A "record" is a written document that organizes past events and data.

[0393] A "terminology glossary" is a dictionary or list of technical terms related to a particular job or field.

[0394] A "content recommendation system" is a system that suggests appropriate content based on a user's specific state or preferences.

[0395] "Collaborative filtering" is a method for recommending new content based on a user's preferences and behavior.

[0396] "Content-based filtering" is a method of recommending suitable content to a user based on the characteristics of the content itself.

[0397] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[0398] User operations

[0399] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a medical translation AI, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[0400] Device behavior

[0401] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0402] Server Operation

[0403] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[0404] Training an AI model

[0405] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0406] How the recommendation system works

[0407] Furthermore, the server also has a content recommendation system based on the user's emotional state. For example, if the user's emotional state is estimated to be a desire to relax, it will recommend relaxing content. The recommendation algorithm uses collaborative filtering and content-based filtering.

[0408] Training results storage and notification

[0409] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[0410] Displaying results on a terminal

[0411] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[0412] Specific examples

[0413] For example, the following prompts can be input to a generative AI model:

[0414] Prompt Sentence Examples

[0415] It is assumed that the user's current emotional state is that they want to relax. Please recommend three relaxing video or music content.

[0416] This allows the system to efficiently provide relaxing content to the user and provide an optimal content viewing experience.

[0417] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0418] Step 1:

[0419] The user enters the "Industry" and "Model Type" through the interface and clicks the "Train AI" button.

[0420] Input: User-entered industry (e.g., "Medical") and model type (e.g., "translator").

[0421] Output: Data formatted in JSON format.

[0422] What happens: The user initiates a training request by entering the required information on the device screen.

[0423] Step 2:

[0424] The device collects the user's input data and the emotional state obtained from the emotion engine, and generates JSON data.

[0425] Inputs: User-entered industry, model type, and sentiment state from the sentiment engine.

[0426] Output: JSON data to send to the server.

[0427] Specific operation: The device combines the input information and the emotion data obtained from the emotion engine into a single data format (JSON).

[0428] Step 3:

[0429] The device sends the generated JSON data to the server's / train-ai endpoint as a POST request.

[0430] Input: Data in JSON format.

[0431] Output: Data is sent to the server.

[0432] What it does: Sends formatted data to the server using a POST request.

[0433] Step 4:

[0434] The server parses the received JSON data and extracts the included "Industry," "Domain Data," "Model Type," and "Emotional State."

[0435] Input: JSON data sent from the terminal.

[0436] Output: Data for each parsed field.

[0437] Specific operation: The server analyzes the data and extracts the necessary information individually.

[0438] Step 5:

[0439] The server starts training the AI ​​model based on the received data.

[0440] Input: Parsed "Industry", "Domain Data", "Model Type", "Emotional State".

[0441] Output: The trained AI model.

[0442] Specific operation: The server selects the specified neural network model and adjusts the training parameters based on the user's emotional state.

[0443] Step 6:

[0444] The server preprocesses the Domain Data and converts it into a trainable format.

[0445] Input: Raw "Domain Data".

[0446] Output: Preprocessed data.

[0447] Specific operation: Performs preprocessing such as data cleaning and normalization to prepare the data in a format that can be used by the model.

[0448] Step 7:

[0449] Save the trained AI model and generate a unique identifier (model ID).

[0450] Input: A trained AI model.

[0451] Output: Model ID.

[0452] Specific operation: The trained AI model is saved in a database and an ID is generated to identify the model.

[0453] Step 8:

[0454] The server generates a JSON response containing a training result success message and the model ID and sends it to the device.

[0455] Input: Training result, model ID.

[0456] Output: JSON response.

[0457] Specific operation: The server organizes the obtained model ID and training results, generates a response message, and sends it to the terminal.

[0458] Step 9:

[0459] The terminal receives the response from the server, analyzes the content, and displays it to the user.

[0460] Input: JSON response from the server.

[0461] Output: A success message and the model ID.

[0462] What it does: The device converts the received message into a user-friendly format and displays it on the screen, for example, "Training successful. Model ID: model12345."

[0463] 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.

[0464] 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.

[0465] 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.

[0466] [Second embodiment]

[0467] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0468] 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.

[0469] 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).

[0470] 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.

[0471] 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.

[0472] 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).

[0473] 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.

[0474] 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.

[0475] 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.

[0476] 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.

[0477] 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.

[0478] 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."

[0479] This invention relates to a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The operation of the system is explained below.

[0480] User operations

[0481] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a translation AI in the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, users can click the "Train AI" button to start the AI ​​training request.

[0482] Device behavior

[0483] The terminal collects the data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," and "Model Type." The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0484] Server Operation

[0485] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," and "Model Type." The server selects a model for training and starts training the AI ​​model based on the received data.

[0486] Training an AI model

[0487] The server selects a neural network model for training based on the received "Model Type." For example, for a translation model, it configures and applies a neural network with specific input and output sizes. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0488] Training results storage and notification

[0489] Once training is complete, the server stores the trained model along with a unique identifier that will be used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[0490] Displaying results on a terminal

[0491] When the device receives the response from the server, it analyzes the response and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message such as "Training successful. Model ID: model12345" may be displayed.

[0492] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[0493] The processing flow will be explained below.

[0494] Step 1:

[0495] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[0496] Step 2:

[0497] The user clicks the "Train AI" button.

[0498] Step 3:

[0499] The device calls the JavaScript trainAI function.

[0500] Step 4:

[0501] The terminal collects the input data of "Industry" and "Model Type".

[0502] Step 5:

[0503] The data collected by the device is converted into JSON format. For example, it is converted into the following format.

[0504] json

[0505] {

[0506] "industry": "Medical",

[0507] "domain_data": [], / / Include medical related data here

[0508] "model_type": "translator"

[0509] }

[0510] Step 6:

[0511] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[0512] Step 7:

[0513] The server receives the POST request and parses the JSON data.

[0514] Step 8:

[0515] The server extracts the industry, domain_data, and model_type fields from the parsed data.

[0516] Step 9:

[0517] The server calls the train_model function to start training the AI ​​model.

[0518] Step 10:

[0519] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[0520] Step 11:

[0521] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[0522] Step 12:

[0523] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[0524] Step 13:

[0525] After the server completes training, it saves the model to storage.

[0526] Step 14:

[0527] The server assigns a unique identification (model ID) to the saved model.

[0528] Step 15:

[0529] The server generates a JSON response containing a training success message and the model ID.

[0530] json

[0531] {

[0532] "message": "Training successful",

[0533] "model_id": "model12345"

[0534] }

[0535] Step 16:

[0536] The server generates a JSON response and sends it back to the device.

[0537] Step 17:

[0538] The device receives a JSON response from the server.

[0539] Step 18:

[0540] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[0541] Example 1

[0542] 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."

[0543] An appropriate training process is necessary for general-purpose AI to effectively learn specialized knowledge and skills for specific tasks and evolve into AI with advanced specialized knowledge. However, conventional methods have not been able to achieve efficient training because they do not effectively utilize user input data and do not fully reflect expert knowledge and guidance. Furthermore, training results are not returned to users quickly and reliably, making them less convenient to use in actual tasks.

[0544] 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.

[0545] In this invention, the server includes: a means for a user to select a specific job or industry and initiate a training request; a means for a terminal to collect user input data, format it in JSON format, and send it to the server; a means for the server to analyze the received data, select a specific model, and initiate AI training based on the data; a means for saving the AI ​​model after completion of training and generating unique identification information; and a means for providing the user with a training result success message including the identification information. This allows the general AI to effectively learn the specialized knowledge and skills required for a specific job, and makes it possible to quickly and reliably provide the trained AI model to the user.

[0546] "General-purpose artificial intelligence" is artificial intelligence that is not limited to a specific use and can handle a variety of tasks and operations.

[0547] "Specific work" refers to specific tasks or activities related to a particular job, role, or industry.

[0548] "Expertise" refers to specialized knowledge or information related to a particular job or field.

[0549] "Skills" refers to the techniques and abilities required to perform a particular task efficiently and accurately.

[0550] An "expert" is someone who has advanced knowledge and experience in a particular job or field.

[0551] "Training data" refers to the datasets used to train artificial intelligence, including specialized literature, documents, records, glossaries, etc.

[0552] "User" refers to any person or organization that uses the System to request training of the Artificial Intelligence.

[0553] A "terminal" is a device used by a user to operate the system, and includes a computer, a smartphone, etc.

[0554] The "JSON format" is a lightweight data exchange format for structuring and representing data.

[0555] A "server" refers to a computer system that receives and processes requests from users.

[0556] "Analysis" refers to the process of understanding the data received and extracting the necessary information.

[0557] A "neural network model" is a specific algorithmic model used to train artificial intelligence, which is based on neural networks.

[0558] "Preprocessing" refers to the process for converting training data into a trainable form.

[0559] "Training" refers to the process of teaching an artificial intelligence model the ability to handle a specific task.

[0560] "Identification information" refers to information that uniquely identifies a trained artificial intelligence model.

[0561] A "success message" refers to a message that informs the user that the training is complete.

[0562] The present invention is a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The detailed implementation method of the system is described below.

[0563] First, the user accesses the system's operation screen and enters data to select a specific business or industry. Specifically, the user enters the industry name in the "Industry" field and the model type in the "Model Type" field. For example, if the user wants to train a translation AI for the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to start the training request.

[0564] Next, the terminal collects the data entered by the user and formats it in JSON format. For example, it generates JSON data containing information such as "Industry," "Domain Data," and "Model Type." This data is sent from the terminal to the specified endpoint of the server as a POST request. The terminal automatically formats the data and sends it to the server.

[0565] The server receives the request sent from the terminal and analyzes the included JSON data, which includes "Industry," "Domain Data," and "Model Type." Based on the analysis results, the server selects a model for training and begins training the AI ​​model based on the data. Specifically, the server selects a neural network model according to the "Model Type" and preprocesses the received "Domain Data" to convert it into a trainable format.

[0566] The server then uses the preprocessed data to train the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. Through this process, the AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[0567] Once training is complete, the server stores the trained model along with a unique identifier, which is used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[0568] Finally, the device receives the response from the server, analyzes it, and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message like "Training successful. Model ID: model12345" can be displayed.

[0569] Take the training of medical translation AI as a concrete example. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[0570] Prompt Sentence Examples

[0571] Prompt: I have created a dataset to train a translation AI specialized for the medical field. Now, please train your AI model on this dataset.

[0572] Input data: "Industry: Medical" "Domain Data: Medical books, papers, medical records, terminology" "Model Type: translator"

[0573] Output: The AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[0574] By linking user input with server processing, this system effectively specializes artificial intelligence for specific tasks, allowing users to quickly and efficiently utilize AI models with the necessary expertise and skills.

[0575] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0576] Step 1:

[0577] The user enters data

[0578] Specific behavior:

[0579] The user accesses the system's operation screen and enters data into the fields for selecting the job and industry. Specifically, enter the industry name in the "Industry" field and the model type in the "Model Type" field. For example, enter "Medical" in "Industry" and "Translator" in "Model Type." Then, click the "Train AI" button to start the training request.

[0580] Input: User-entered industry name and model type

[0581] Output: Training request started

[0582] Step 2:

[0583] The device collects and formats the data

[0584] Specific behavior:

[0585] The terminal collects the "Industry," "Domain Data," and "Model Type" data entered by the user. It formats the collected data in JSON format and prepares it for sending to the server. The generated JSON data looks like this, for example:

[0586] {

[0587] "Industry": "Medical",

[0588] "Domain Data": "Medical books, papers, medical records, terminology",

[0589] "Model Type": "translator"

[0590] }

[0591] Input: User-entered industry name, domain data, and model type

[0592] Output: JSON format data

[0593] Step 3:

[0594] The device sends data to the server

[0595] Specific behavior:

[0596] The formatted JSON data is sent as a POST request to the server's / train-ai endpoint. The device checks whether the transmission was successful.

[0597] Input: JSON format data

[0598] Output: Sending a POST request to the server

[0599] Step 4:

[0600] The server analyzes the received data

[0601] Specific behavior:

[0602] The server receives the request sent from the device and parses the included JSON data, extracting the following information from the parsed data: "Industry," "Domain Data," and "Model Type."

[0603] Input: Received JSON data

[0604] Output: Parsed "Industry", "Domain Data", and "Model Type"

[0605] Step 5:

[0606] The server selects the training model and preprocesses the data.

[0607] Specific behavior:

[0608] The server selects an appropriate neural network model based on the extracted "Model Type," and then preprocesses the acquired "Domain Data" and converts it into a trainable format.

[0609] Input: Parsed "Model Type" and "Domain Data"

[0610] Output: Data converted into a trainable format

[0611] Step 6:

[0612] The server trains the AI ​​model

[0613] Specific behavior:

[0614] Using the preprocessed data, the server trains the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. In this process, the AI ​​model acquires skills specific to medical translation.

[0615] Input: Data converted into a trainable format

[0616] Output: A trained AI model

[0617] Step 7:

[0618] The server stores and notifies the training results

[0619] Specific behavior:

[0620] Once training is complete, the server saves the model and generates a unique identifier (Model ID). The server generates a response containing a training success message and the unique identifier and sends it to the device.

[0621] Input: A trained AI model

[0622] Output: Model identification and a success message

[0623] Step 8:

[0624] The device receives the server's response and displays it to the user.

[0625] Specific behavior:

[0626] The device receives the response from the server, analyzes its contents, and displays a message indicating successful training and the model's identification information to the user. For example, a message such as "Training successful. Model ID: model12345" is displayed.

[0627] Input: Response from the server

[0628] Output: Displaying training results to the user

[0629] (Application example 1)

[0630] 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."

[0631] In modern brick-and-mortar stores, it is difficult for customers to quickly obtain detailed product information and usage instructions. Furthermore, there is a lack of a system that properly recommends related products, which prevents customers from making optimal choices when purchasing. Furthermore, inventory cannot be checked in real time, which causes inconvenience for both customers and stores.

[0632] 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.

[0633] In this invention, the server includes: means for introducing a general-purpose artificial intelligence and learning specialized knowledge and skills for a specific job; means for evolving the general-purpose artificial intelligence into an artificial intelligence suitable for the specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence; means for recommending detailed information, usage guides, and related products related to the specific job; and means for reading barcodes to obtain product information. This allows customers to obtain product information, usage instructions, and related product recommendations in real time in stores, improving the accuracy of their purchasing decisions.

[0634] "General-purpose artificial intelligence" is artificial intelligence that is not dependent on specific tasks or jobs and can be applied in a wide range of fields.

[0635] "Specialized knowledge" is detailed and advanced knowledge acquired in a particular job or industry.

[0636] "Technology" refers to the specific techniques and skills required for a particular job or industry.

[0637] An "industry expert" is a professional with extensive experience and in-depth knowledge in a particular industry.

[0638] A "specific job" is a specific job or role performed in a particular industry or field.

[0639] A "specialty book" is a publication that provides advanced knowledge or information in a specific field.

[0640] A "paper" is a document that reports the results of detailed research and analysis on a specific research subject.

[0641] A "record" is information that preserves past facts, events, and data.

[0642] A "terminology glossary" is a dictionary or list of technical terms in a particular field.

[0643] "Detailed information" means specific and comprehensive information about a product or service.

[0644] A "usage guide" is a document that explains the specific steps and methods for using a product or service.

[0645] "Related products" are other products that are highly related to a particular product.

[0646] "Recommendation" is the act of introducing something to others based on specific conditions or requirements.

[0647] A "barcode" is a series of lines and numbers used to represent the identity of a product or item.

[0648] "Product information" is detailed data about the product's characteristics, specifications, usage, price, etc.

[0649] "Stock status" is information that indicates how much of a particular product is currently in stock.

[0650] This invention provides a system that allows customers in physical stores to quickly obtain detailed product information, usage instructions, and related product recommendations. This system employs general artificial intelligence (AI) and is trained to acquire specialized knowledge and skills corresponding to specific tasks.

[0651] Hardware and Software Configuration

[0652] The system's hardware consists of a smartphone, a server, and a barcode reader, while the software uses an API client developed in Python, TensorFlow or PyTorch for training the neural network model, and Flask or Django for providing the REST API server.

[0653] Feature details

[0654] 1. Training the general-purpose AI: The server uses specialized books, papers, records, and terminology to train the general-purpose AI with specialized knowledge and skills for specific tasks. This data is sent to the server in JSON format and used to train the AI ​​model.

[0655] 2. Product information acquisition: Using a smartphone application, the customer scans the product barcode, which retrieves the corresponding product information from the server and displays it on the smartphone screen.

[0656] 3. Providing usage guide: Based on the product ID, the server provides instructions on how to use and assemble the corresponding product. This information is also displayed through the smartphone application.

[0657] 4. Related product recommendation: The server recommends other related products based on the specific product information, and this recommendation information is also displayed on the smartphone screen.

[0658] 5. Check stock availability: The server can check the stock availability of a particular product in real time and provide that information on the smartphone screen.

[0659] Processing flow

[0660] User operation: The user operates the smartphone app and scans the barcode.

[0661] Terminal operation: Collects barcode information and sends it to the server in JSON format.

[0662] Server Action: Parse the product information based on the received barcode information and generate and return a response containing detailed information, usage guides, related product recommendations, and stock availability.

[0663] Displaying the results: The smartphone app receives the response from the server and displays it to the user.

[0664] Specific examples

[0665] For example, if a user wants to get information about the latest running shoes, they can scan the barcode with their smartphone and instantly see detailed product information, instructions for use, recommendations for related running gear, and the product's availability.

[0666] Prompt Sentence Examples

[0667] "I'm looking for the latest running shoes. I want something durable and suitable for long distance running. What are the best options?"

[0668] In this way, an intelligent system that provides advanced support for the shopping experience in physical stores will be realized.

[0669] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0670] Step 1:

[0671] A user launches a smartphone application and scans a barcode. The barcode reader is used to obtain the product's barcode information, which is then entered into the application. This barcode information is a series of numbers and letters that identify a specific product.

[0672] Step 2:

[0673] The device (smartphone) formats the acquired barcode information into JSON format. This data includes the barcode information and the user's device identification information. An HTTP POST request is generated to send the formatted JSON data to the server.

[0674] Step 3:

[0675] The server receives the JSON data sent from the device. The received data includes the product barcode information. The server analyzes this information and retrieves the corresponding product details from the database.

[0676] Step 4:

[0677] The server generates usage guides and related product data based on the acquired product details. Product usage guides are data that includes procedures and operating methods, and related product data is information on other recommended products. These data are processed on the server and converted into JSON format.

[0678] Step 5:

[0679] The server queries the inventory management system to check the real-time availability of a product. The query returns the product's stock quantity and its status. The server also converts this stock information into JSON format.

[0680] Step 6:

[0681] The server compiles the generated product details, usage guide, related product information, and stock status into a single JSON response and sends this response to the device as an HTTP response. The response contains all the necessary product information.

[0682] Step 7:

[0683] The device parses the JSON response received from the server and displays it on the screen. Users can then use the smartphone application to check product details, usage guides, related product recommendations, and stock availability, allowing them to make product selections.

[0684] These steps enable users to enjoy an efficient and high-quality shopping experience in physical stores. For example, when selecting running shoes, users can obtain optimal product information by entering a prompt such as, "I'm looking for the latest running shoes. I want something durable and suitable for long-distance running. Which product would be best?"

[0685] 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.

[0686] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[0687] User operations

[0688] Users first select the specific job or industry they want to give expertise in, entering "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, users click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[0689] Device behavior

[0690] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0691] Server Operation

[0692] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[0693] Training an AI model

[0694] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0695] Training results storage and notification

[0696] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[0697] Displaying results on a terminal

[0698] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[0699] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis certificate, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[0700] The processing flow will be explained below.

[0701] Step 1:

[0702] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[0703] Step 2:

[0704] The user clicks the "Train AI" button.

[0705] Step 3:

[0706] The device calls the JavaScript trainAI function.

[0707] Step 4:

[0708] The terminal collects the input data of "Industry" and "Model Type".

[0709] Step 5:

[0710] The device uses an emotion engine to detect the user's emotional state in real time, for example by analyzing emotions from the user's facial expressions and tone of voice.

[0711] Step 6:

[0712] The device converts the collected data and emotional state into a JSON format, for example, as follows:

[0713] json

[0714] {

[0715] "industry": "Medical",

[0716] "domain_data": [], / / Include medical related data here

[0717] "model_type": "translator",

[0718] "emotional_state": "neutral" / / or "stressed", etc.

[0719] }

[0720] Step 7:

[0721] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[0722] Step 8:

[0723] The server receives the POST request and parses the JSON data.

[0724] Step 9:

[0725] The server extracts the industry, domain_data, model_type, and emotional_state fields from the parsed data.

[0726] Step 10:

[0727] The server calls the train_model function to start training the AI ​​model.

[0728] Step 11:

[0729] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[0730] Step 12:

[0731] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[0732] Step 13:

[0733] The server adjusts the training parameters based on the user's emotional state (emotional_state). For example, if the user is feeling stressed, the difficulty of the training will be reduced.

[0734] Step 14:

[0735] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[0736] Step 15:

[0737] After the server completes training, it saves the model to storage.

[0738] Step 16:

[0739] The server assigns a unique identification (model ID) to the saved model.

[0740] Step 17:

[0741] The server generates a JSON response containing a training success message and the model ID.

[0742] json

[0743] {

[0744] "message": "Training successful",

[0745] "model_id": "model12345"

[0746] }

[0747] Step 18:

[0748] The server generates a JSON response and sends it back to the device.

[0749] Step 19:

[0750] The device receives a JSON response from the server.

[0751] Step 20:

[0752] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[0753] Example 2

[0754] 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."

[0755] Conventional general-purpose AI systems have difficulty effectively learning specialized knowledge and skills specific to specific jobs or industries. Furthermore, because they do not take the user's emotional state into account, the effectiveness of training depends on the user's psychological state, resulting in reduced efficiency. Therefore, a new system that can monitor the user's emotional state in real time and optimize training is needed.

[0756] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: means for learning specialized knowledge and skills for a specific job; means for evolving a general-purpose AI into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as training data; means for monitoring the user's emotional state in real time; means for collecting data entered by the user and formatting the data in JSON format; means for transmitting the formatted data to the server; and means for analyzing the received data and starting AI training based on the data. This enables efficient training of AI specialized for a specific job, and optimal training that takes into account the user's emotional state can be achieved.

[0757] "General-purpose artificial intelligence" is a general-purpose artificial intelligence system that can be used in a wide range of areas, without being limited to specific tasks or applications.

[0758] A "job" refers to a specific task or role that a person performs in their occupation.

[0759] "Skills" is a general term for the techniques and abilities required to perform specific jobs or tasks.

[0760] An "industry expert" refers to an expert with advanced knowledge and experience in a particular industry or field.

[0761] A "specialized book" is a book that provides specialized knowledge related to a particular field or job.

[0762] A "paper" is an academic piece of writing that describes the results of a particular study or investigation.

[0763] "Records" refer to documents or data that record specific facts or events.

[0764] A "terminology glossary" is a collection of technical terms and their definitions used in a particular field or job.

[0765] An "emotion engine" refers to a system or algorithm that monitors and analyzes a user's emotional state in real time.

[0766] "Real-time monitoring" means continuously observing the user's condition in real time and immediately detecting any changes.

[0767] "Data collection" is the act or process of systematically compiling necessary information.

[0768] "JSON format" is an abbreviation for JavaScript Object Notation, and is a lightweight data exchange format for expressing data in text format.

[0769] A "server" is a computer system that receives and responds to requests from clients on a network.

[0770] A "neural network model" is a type of artificial intelligence with a structure that mimics biological nerve cells, and is an algorithm that specializes in learning and prediction for specific tasks.

[0771] "Training" is the process of providing data to an artificial intelligence or machine learning model to learn and improve its predictive capabilities and performance.

[0772] "Training parameters" refer to settings or values ​​that are adjusted during the learning process of an artificial intelligence model.

[0773] "Preprocessing" refers to the preliminary processing of raw data to convert it into a format that is easy for machine learning models to handle.

[0774] "Identification information" refers to information or an ID that uniquely identifies a specific object or data.

[0775] A "prompt" is a document or instruction used as input to a generative artificial intelligence model.

[0776] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills for a specific job, and also combines it with an emotion engine that recognizes the user's emotions. The system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[0777] Users first select the specific job or industry they want to provide expertise in and enter "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, they click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine, which analyzes the user's emotional state and dynamically adjusts training parameters.

[0778] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0779] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[0780] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0781] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[0782] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[0783] Take the training of medical translation AI as a concrete example. When users train a medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[0784] Example prompt sentence:

[0785] "Train a medical translator AI model using data from medical books, research papers, medical records, and medical glossaries. Adjust training parameters based on user's emotional state, ensuring lower difficulty when the user feels stressed."

[0786] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0787] Step 1: User Input

[0788] Users use the system's interface to select the specific job or industry they want to train in by entering "Medical" in the "Industry" field and "translator" in the "Model Type" field. The emotion engine monitors the user's emotional state in real time and collects that data. The input data consists of "Industry," "Model Type," and "Emotional State." For example, if the user is stressed, the "Emotional State" is recorded as "stressed."

[0789] Step 2: Data formatting and transmission by the terminal

[0790] The device collects the "Industry" and "Model Type" input by the user, as well as the "Emotional State" information obtained from the emotion engine, and formats it in JSON format. For example, the following JSON data is generated:

[0791] json

[0792] {

[0793] "Industry": "Medical",

[0794] "Model Type": "translator",

[0795] "Emotional State": "stressed"

[0796] }

[0797] This JSON data is sent as a POST request to the server's / train-ai endpoint. Data formatting and transmission are automated by the device.

[0798] Step 3: Data reception and analysis by the server

[0799] The server receives the POST request sent from the device and parses the included JSON data. The parsed results include "Industry," "Model Type," and "Emotional State." The server first parses the data and obtains the values ​​of each field. For example, "Industry" is "Medical," "Model Type" is "translator," and "Emotional State" is "stressed."

[0800] Step 4: Selecting an AI model and adjusting its parameters

[0801] The server selects an appropriate neural network model for training based on the analyzed data. In this case, because the "Model Type" is "translator," a model specialized for translation is selected. The server also references the "Emotional State" obtained from the emotion engine and adjusts parameters, such as lowering the difficulty of training if the user is feeling stressed.

[0802] Step 5: Preprocess the data and run training

[0803] The server preprocesses the domain data and converts it into a trainable format. For example, it cleans and formats data from medical papers and diagnostic reports. This preprocessed data is then used to train the AI ​​model. Specifically, the data is fed into a neural network, which learns it epoch by epoch.

[0804] Step 6: Save and notify training results

[0805] Once training is complete, the server saves the trained model with a unique identifier (model ID). For example, the trained model is assigned an ID of "model12345". The server then generates a JSON response containing a training success message and the model's identifier, and sends it to the device. An example response is as follows:

[0806] json

[0807] {

[0808] "message": "Training successful",

[0809] "modelID": "model12345"

[0810] }

[0811] Step 7: Displaying the results on your device

[0812] The device receives the response from the server and analyzes its contents. The analysis results include a success message and a model ID. The device displays this information to the user. For example, it might display "Training successful. Model ID: model12345." By checking this result, the user can confirm that the training was successful and obtain the model's identification information.

[0813] (Application example 2)

[0814] 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."

[0815] Conventional general-purpose AI training systems have had the problem of being difficult to efficiently teach specialized knowledge and skills specialized for specific jobs. Furthermore, the lack of feedback based on the user's emotional state can sometimes prevent the training from being fully effective. Furthermore, there was no system that could recommend appropriate content taking the user's emotional state into account, resulting in a less than satisfactory user experience.

[0816] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0817] In this invention, the server includes: a means for introducing a general-purpose artificial intelligence (AI) to learn specialized knowledge and skills for a specific job; a means for evolving the AI ​​into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; a means for using specialized books, papers, records, and technical glossaries as training data for the AI; a means for incorporating an emotion engine that analyzes a user's emotional state in real time and provides appropriate feedback according to that state; and a means for generating prompts and adjusting the training of the AI ​​model based on the emotional state. This enables the AI ​​to efficiently learn specialized knowledge and skills specific to a specific job, and provides feedback according to the emotional state to improve the effectiveness of training. Furthermore, the user experience can be improved by recommending content based on the user's emotional state.

[0818] "General artificial intelligence" is artificial intelligence that can handle a wide range of jobs and tasks.

[0819] "Expertise" refers to detailed information or knowledge about a particular job or field.

[0820] "Skills" refer to the techniques and abilities required to perform a particular job.

[0821] An "industry expert" is a person or group of people who have a high level of expertise in a particular job function or field.

[0822] An "emotion engine" is a software or hardware system for detecting and analyzing a user's emotional state.

[0823] "Real-time" refers to a timeframe in which processing and feedback occurs almost immediately.

[0824] A "prompt" is input data used to give instructions or ask questions to artificial intelligence.

[0825] "Training parameters" are settings or values ​​that can be adjusted during the learning process of an artificial intelligence model.

[0826] A "specialized book" is a book that contains detailed information about a specific job or field.

[0827] A "paper" is an academic document that summarizes research results and considerations on a specific topic.

[0828] A "record" is a written document that organizes past events and data.

[0829] A "terminology glossary" is a dictionary or list of technical terms related to a particular job or field.

[0830] A "content recommendation system" is a system that suggests appropriate content based on a user's specific state or preferences.

[0831] "Collaborative filtering" is a method for recommending new content based on a user's preferences and behavior.

[0832] "Content-based filtering" is a method of recommending suitable content to a user based on the characteristics of the content itself.

[0833] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[0834] User operations

[0835] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a medical translation AI, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[0836] Device behavior

[0837] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0838] Server Operation

[0839] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[0840] Training an AI model

[0841] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0842] How the recommendation system works

[0843] Furthermore, the server also has a content recommendation system based on the user's emotional state. For example, if the user's emotional state is estimated to be a desire to relax, it will recommend relaxing content. The recommendation algorithm uses collaborative filtering and content-based filtering.

[0844] Training results storage and notification

[0845] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[0846] Displaying results on a terminal

[0847] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[0848] Specific examples

[0849] For example, the following prompts can be input to a generative AI model:

[0850] Prompt Sentence Examples

[0851] It is assumed that the user's current emotional state is that they want to relax. Please recommend three relaxing video or music content.

[0852] This allows the system to efficiently provide relaxing content to the user and provide an optimal content viewing experience.

[0853] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0854] Step 1:

[0855] The user enters the "Industry" and "Model Type" through the interface and clicks the "Train AI" button.

[0856] Input: User-entered industry (e.g., "Medical") and model type (e.g., "translator").

[0857] Output: Data formatted in JSON format.

[0858] What happens: The user initiates a training request by entering the required information on the device screen.

[0859] Step 2:

[0860] The device collects the user's input data and the emotional state obtained from the emotion engine, and generates JSON data.

[0861] Inputs: User-entered industry, model type, and sentiment state from the sentiment engine.

[0862] Output: JSON data to send to the server.

[0863] Specific operation: The device combines the input information and the emotion data obtained from the emotion engine into a single data format (JSON).

[0864] Step 3:

[0865] The device sends the generated JSON data to the server's / train-ai endpoint as a POST request.

[0866] Input: Data in JSON format.

[0867] Output: Data is sent to the server.

[0868] What it does: Sends formatted data to the server using a POST request.

[0869] Step 4:

[0870] The server parses the received JSON data and extracts the included "Industry," "Domain Data," "Model Type," and "Emotional State."

[0871] Input: JSON data sent from the terminal.

[0872] Output: Data for each parsed field.

[0873] Specific operation: The server analyzes the data and extracts the necessary information individually.

[0874] Step 5:

[0875] The server starts training the AI ​​model based on the received data.

[0876] Input: Parsed "Industry", "Domain Data", "Model Type", "Emotional State".

[0877] Output: The trained AI model.

[0878] Specific operation: The server selects the specified neural network model and adjusts the training parameters based on the user's emotional state.

[0879] Step 6:

[0880] The server preprocesses the Domain Data and converts it into a trainable format.

[0881] Input: Raw "Domain Data".

[0882] Output: Preprocessed data.

[0883] Specific operation: Performs preprocessing such as data cleaning and normalization to prepare the data in a format that can be used by the model.

[0884] Step 7:

[0885] Save the trained AI model and generate a unique identifier (model ID).

[0886] Input: A trained AI model.

[0887] Output: Model ID.

[0888] Specific operation: The trained AI model is saved in a database and an ID is generated to identify the model.

[0889] Step 8:

[0890] The server generates a JSON response containing a training result success message and the model ID and sends it to the device.

[0891] Input: Training result, model ID.

[0892] Output: JSON response.

[0893] Specific operation: The server organizes the obtained model ID and training results, generates a response message, and sends it to the terminal.

[0894] Step 9:

[0895] The terminal receives the response from the server, analyzes the content, and displays it to the user.

[0896] Input: JSON response from the server.

[0897] Output: A success message and the model ID.

[0898] What it does: The device converts the received message into a user-friendly format and displays it on the screen, for example, "Training successful. Model ID: model12345."

[0899] 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.

[0900] 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.

[0901] 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.

[0902] [Third embodiment]

[0903] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0904] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0905] 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).

[0906] 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.

[0907] 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.

[0908] 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).

[0909] 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.

[0910] 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.

[0911] 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.

[0912] 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.

[0913] 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.

[0914] 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."

[0915] This invention relates to a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The operation of the system is explained below.

[0916] User operations

[0917] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a translation AI in the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, users can click the "Train AI" button to start the AI ​​training request.

[0918] Device behavior

[0919] The terminal collects the data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," and "Model Type." The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[0920] Server Operation

[0921] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," and "Model Type." The server selects a model for training and starts training the AI ​​model based on the received data.

[0922] Training an AI model

[0923] The server selects a neural network model for training based on the received "Model Type." For example, for a translation model, it configures and applies a neural network with specific input and output sizes. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[0924] Training results storage and notification

[0925] Once training is complete, the server stores the trained model along with a unique identifier that will be used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[0926] Displaying results on a terminal

[0927] When the device receives the response from the server, it analyzes the response and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message such as "Training successful. Model ID: model12345" may be displayed.

[0928] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[0929] The processing flow will be explained below.

[0930] Step 1:

[0931] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[0932] Step 2:

[0933] The user clicks the "Train AI" button.

[0934] Step 3:

[0935] The device calls the JavaScript trainAI function.

[0936] Step 4:

[0937] The terminal collects the input data of "Industry" and "Model Type".

[0938] Step 5:

[0939] The data collected by the device is converted into JSON format. For example, it is converted into the following format.

[0940] json

[0941] {

[0942] "industry": "Medical",

[0943] "domain_data": [], / / Include medical related data here

[0944] "model_type": "translator"

[0945] }

[0946] Step 6:

[0947] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[0948] Step 7:

[0949] The server receives the POST request and parses the JSON data.

[0950] Step 8:

[0951] The server extracts the industry, domain_data, and model_type fields from the parsed data.

[0952] Step 9:

[0953] The server calls the train_model function to start training the AI ​​model.

[0954] Step 10:

[0955] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[0956] Step 11:

[0957] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[0958] Step 12:

[0959] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[0960] Step 13:

[0961] After the server completes training, it saves the model to storage.

[0962] Step 14:

[0963] The server assigns a unique identification (model ID) to the saved model.

[0964] Step 15:

[0965] The server generates a JSON response containing a training success message and the model ID.

[0966] json

[0967] {

[0968] "message": "Training successful",

[0969] "model_id": "model12345"

[0970] }

[0971] Step 16:

[0972] The server generates a JSON response and sends it back to the device.

[0973] Step 17:

[0974] The device receives a JSON response from the server.

[0975] Step 18:

[0976] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[0977] Example 1

[0978] 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."

[0979] An appropriate training process is necessary for general-purpose AI to effectively learn specialized knowledge and skills for specific tasks and evolve into AI with advanced specialized knowledge. However, conventional methods have not been able to achieve efficient training because they do not effectively utilize user input data and do not fully reflect expert knowledge and guidance. Furthermore, training results are not returned to users quickly and reliably, making them less convenient to use in actual tasks.

[0980] 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.

[0981] In this invention, the server includes: a means for a user to select a specific job or industry and initiate a training request; a means for a terminal to collect user input data, format it in JSON format, and send it to the server; a means for the server to analyze the received data, select a specific model, and initiate AI training based on the data; a means for saving the AI ​​model after completion of training and generating unique identification information; and a means for providing the user with a training result success message including the identification information. This allows the general AI to effectively learn the specialized knowledge and skills required for a specific job, and makes it possible to quickly and reliably provide the trained AI model to the user.

[0982] "General-purpose artificial intelligence" is artificial intelligence that is not limited to a specific use and can handle a variety of tasks and operations.

[0983] "Specific work" refers to specific tasks or activities related to a particular job, role, or industry.

[0984] "Expertise" refers to specialized knowledge or information related to a particular job or field.

[0985] "Skills" refers to the techniques and abilities required to perform a particular task efficiently and accurately.

[0986] An "expert" is someone who has advanced knowledge and experience in a particular job or field.

[0987] "Training data" refers to the datasets used to train artificial intelligence, including specialized literature, documents, records, glossaries, etc.

[0988] "User" refers to any person or organization that uses the System to request training of the Artificial Intelligence.

[0989] A "terminal" is a device used by a user to operate the system, and includes a computer, a smartphone, etc.

[0990] The "JSON format" is a lightweight data exchange format for structuring and representing data.

[0991] A "server" refers to a computer system that receives and processes requests from users.

[0992] "Analysis" refers to the process of understanding the data received and extracting the necessary information.

[0993] A "neural network model" is a specific algorithmic model used to train artificial intelligence, which is based on neural networks.

[0994] "Preprocessing" refers to the process for converting training data into a trainable form.

[0995] "Training" refers to the process of teaching an artificial intelligence model the ability to handle a specific task.

[0996] "Identification information" refers to information that uniquely identifies a trained artificial intelligence model.

[0997] A "success message" refers to a message that informs the user that the training is complete.

[0998] The present invention is a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The detailed implementation method of the system is described below.

[0999] First, the user accesses the system's operation screen and enters data to select a specific business or industry. Specifically, the user enters the industry name in the "Industry" field and the model type in the "Model Type" field. For example, if the user wants to train a translation AI for the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to start the training request.

[1000] Next, the terminal collects the data entered by the user and formats it in JSON format. For example, it generates JSON data containing information such as "Industry," "Domain Data," and "Model Type." This data is sent from the terminal to the specified endpoint of the server as a POST request. The terminal automatically formats the data and sends it to the server.

[1001] The server receives the request sent from the terminal and analyzes the included JSON data, which includes "Industry," "Domain Data," and "Model Type." Based on the analysis results, the server selects a model for training and begins training the AI ​​model based on the data. Specifically, the server selects a neural network model according to the "Model Type" and preprocesses the received "Domain Data" to convert it into a trainable format.

[1002] The server then uses the preprocessed data to train the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. Through this process, the AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[1003] Once training is complete, the server stores the trained model along with a unique identifier, which is used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[1004] Finally, the device receives the response from the server, analyzes it, and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message like "Training successful. Model ID: model12345" can be displayed.

[1005] Take the training of medical translation AI as a concrete example. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[1006] Prompt Sentence Examples

[1007] Prompt: I have created a dataset to train a translation AI specialized for the medical field. Now, please train your AI model on this dataset.

[1008] Input data: "Industry: Medical" "Domain Data: Medical books, papers, medical records, terminology" "Model Type: translator"

[1009] Output: The AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[1010] By linking user input with server processing, this system effectively specializes artificial intelligence for specific tasks, allowing users to quickly and efficiently utilize AI models with the necessary expertise and skills.

[1011] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1012] Step 1:

[1013] The user enters data

[1014] Specific behavior:

[1015] The user accesses the system's operation screen and enters data into the fields for selecting the job and industry. Specifically, enter the industry name in the "Industry" field and the model type in the "Model Type" field. For example, enter "Medical" in "Industry" and "Translator" in "Model Type." Then, click the "Train AI" button to start the training request.

[1016] Input: User-entered industry name and model type

[1017] Output: Training request started

[1018] Step 2:

[1019] The device collects and formats the data

[1020] Specific behavior:

[1021] The terminal collects the "Industry," "Domain Data," and "Model Type" data entered by the user. It formats the collected data in JSON format and prepares it for sending to the server. The generated JSON data looks like this, for example:

[1022] {

[1023] "Industry": "Medical",

[1024] "Domain Data": "Medical books, papers, medical records, terminology",

[1025] "Model Type": "translator"

[1026] }

[1027] Input: User-entered industry name, domain data, and model type

[1028] Output: JSON format data

[1029] Step 3:

[1030] The device sends data to the server

[1031] Specific behavior:

[1032] The formatted JSON data is sent as a POST request to the server's / train-ai endpoint. The device checks whether the transmission was successful.

[1033] Input: JSON format data

[1034] Output: Sending a POST request to the server

[1035] Step 4:

[1036] The server analyzes the received data

[1037] Specific behavior:

[1038] The server receives the request sent from the device and parses the included JSON data, extracting the following information from the parsed data: "Industry," "Domain Data," and "Model Type."

[1039] Input: Received JSON data

[1040] Output: Parsed "Industry", "Domain Data", and "Model Type"

[1041] Step 5:

[1042] The server selects the training model and preprocesses the data.

[1043] Specific behavior:

[1044] The server selects an appropriate neural network model based on the extracted "Model Type," and then preprocesses the acquired "Domain Data" and converts it into a trainable format.

[1045] Input: Parsed "Model Type" and "Domain Data"

[1046] Output: Data converted into a trainable format

[1047] Step 6:

[1048] The server trains the AI ​​model

[1049] Specific behavior:

[1050] Using the preprocessed data, the server trains the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. In this process, the AI ​​model acquires skills specific to medical translation.

[1051] Input: Data converted into a trainable format

[1052] Output: A trained AI model

[1053] Step 7:

[1054] The server stores and notifies the training results

[1055] Specific behavior:

[1056] Once training is complete, the server saves the model and generates a unique identifier (Model ID). The server generates a response containing a training success message and the unique identifier and sends it to the device.

[1057] Input: A trained AI model

[1058] Output: Model identification and a success message

[1059] Step 8:

[1060] The device receives the server's response and displays it to the user.

[1061] Specific behavior:

[1062] The device receives the response from the server, analyzes its contents, and displays a message indicating successful training and the model's identification information to the user. For example, a message such as "Training successful. Model ID: model12345" is displayed.

[1063] Input: Response from the server

[1064] Output: Displaying training results to the user

[1065] (Application example 1)

[1066] 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."

[1067] In modern brick-and-mortar stores, it is difficult for customers to quickly obtain detailed product information and usage instructions. Furthermore, there is a lack of a system that properly recommends related products, which prevents customers from making optimal choices when purchasing. Furthermore, inventory cannot be checked in real time, which causes inconvenience for both customers and stores.

[1068] 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.

[1069] In this invention, the server includes: means for introducing a general-purpose artificial intelligence and learning specialized knowledge and skills for a specific job; means for evolving the general-purpose artificial intelligence into an artificial intelligence suitable for the specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence; means for recommending detailed information, usage guides, and related products related to the specific job; and means for reading barcodes to obtain product information. This allows customers to obtain product information, usage instructions, and related product recommendations in real time in stores, improving the accuracy of their purchasing decisions.

[1070] "General-purpose artificial intelligence" is artificial intelligence that is not dependent on specific tasks or jobs and can be applied in a wide range of fields.

[1071] "Specialized knowledge" is detailed and advanced knowledge acquired in a particular job or industry.

[1072] "Technology" refers to the specific techniques and skills required for a particular job or industry.

[1073] An "industry expert" is a professional with extensive experience and in-depth knowledge in a particular industry.

[1074] A "specific job" is a specific job or role performed in a particular industry or field.

[1075] A "specialty book" is a publication that provides advanced knowledge or information in a specific field.

[1076] A "paper" is a document that reports the results of detailed research and analysis on a specific research subject.

[1077] A "record" is information that preserves past facts, events, and data.

[1078] A "terminology glossary" is a dictionary or list of technical terms in a particular field.

[1079] "Detailed information" means specific and comprehensive information about a product or service.

[1080] A "usage guide" is a document that explains the specific steps and methods for using a product or service.

[1081] "Related products" are other products that are highly related to a particular product.

[1082] "Recommendation" is the act of introducing something to others based on specific conditions or requirements.

[1083] A "barcode" is a series of lines and numbers used to represent the identity of a product or item.

[1084] "Product information" is detailed data about the product's characteristics, specifications, usage, price, etc.

[1085] "Stock status" is information that indicates how much of a particular product is currently in stock.

[1086] This invention provides a system that allows customers in physical stores to quickly obtain detailed product information, usage instructions, and related product recommendations. This system employs general artificial intelligence (AI) and is trained to acquire specialized knowledge and skills corresponding to specific tasks.

[1087] Hardware and Software Configuration

[1088] The system's hardware consists of a smartphone, a server, and a barcode reader, while the software uses an API client developed in Python, TensorFlow or PyTorch for training the neural network model, and Flask or Django for providing the REST API server.

[1089] Feature details

[1090] 1. Training the general-purpose AI: The server uses specialized books, papers, records, and terminology to train the general-purpose AI with specialized knowledge and skills for specific tasks. This data is sent to the server in JSON format and used to train the AI ​​model.

[1091] 2. Product information acquisition: Using a smartphone application, the customer scans the product barcode, which retrieves the corresponding product information from the server and displays it on the smartphone screen.

[1092] 3. Providing usage guide: Based on the product ID, the server provides instructions on how to use and assemble the corresponding product. This information is also displayed through the smartphone application.

[1093] 4. Related product recommendation: The server recommends other related products based on the specific product information, and this recommendation information is also displayed on the smartphone screen.

[1094] 5. Check stock availability: The server can check the stock availability of a particular product in real time and provide that information on the smartphone screen.

[1095] Processing flow

[1096] User operation: The user operates the smartphone app and scans the barcode.

[1097] Terminal operation: Collects barcode information and sends it to the server in JSON format.

[1098] Server Action: Parse the product information based on the received barcode information and generate and return a response containing detailed information, usage guides, related product recommendations, and stock availability.

[1099] Displaying the results: The smartphone app receives the response from the server and displays it to the user.

[1100] Specific examples

[1101] For example, if a user wants to get information about the latest running shoes, they can scan the barcode with their smartphone and instantly see detailed product information, instructions for use, recommendations for related running gear, and the product's availability.

[1102] Prompt Sentence Examples

[1103] "I'm looking for the latest running shoes. I want something durable and suitable for long distance running. What are the best options?"

[1104] In this way, an intelligent system that provides advanced support for the shopping experience in physical stores will be realized.

[1105] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1106] Step 1:

[1107] A user launches a smartphone application and scans a barcode. The barcode reader is used to obtain the product's barcode information, which is then entered into the application. This barcode information is a series of numbers and letters that identify a specific product.

[1108] Step 2:

[1109] The device (smartphone) formats the acquired barcode information into JSON format. This data includes the barcode information and the user's device identification information. An HTTP POST request is generated to send the formatted JSON data to the server.

[1110] Step 3:

[1111] The server receives the JSON data sent from the device. The received data includes the product barcode information. The server analyzes this information and retrieves the corresponding product details from the database.

[1112] Step 4:

[1113] The server generates usage guides and related product data based on the acquired product details. Product usage guides are data that includes procedures and operating methods, and related product data is information on other recommended products. These data are processed on the server and converted into JSON format.

[1114] Step 5:

[1115] The server queries the inventory management system to check the real-time availability of a product. The query returns the product's stock quantity and its status. The server also converts this stock information into JSON format.

[1116] Step 6:

[1117] The server compiles the generated product details, usage guide, related product information, and stock status into a single JSON response and sends this response to the device as an HTTP response. The response contains all the necessary product information.

[1118] Step 7:

[1119] The device parses the JSON response received from the server and displays it on the screen. Users can then use the smartphone application to check product details, usage guides, related product recommendations, and stock availability, allowing them to make product selections.

[1120] These steps enable users to enjoy an efficient and high-quality shopping experience in physical stores. For example, when selecting running shoes, users can obtain optimal product information by entering a prompt such as, "I'm looking for the latest running shoes. I want something durable and suitable for long-distance running. Which product would be best?"

[1121] 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.

[1122] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[1123] User operations

[1124] Users first select the specific job or industry they want to give expertise in, entering "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, users click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[1125] Device behavior

[1126] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[1127] Server Operation

[1128] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[1129] Training an AI model

[1130] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[1131] Training results storage and notification

[1132] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[1133] Displaying results on a terminal

[1134] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[1135] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis certificate, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[1136] The processing flow will be explained below.

[1137] Step 1:

[1138] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[1139] Step 2:

[1140] The user clicks the "Train AI" button.

[1141] Step 3:

[1142] The device calls the JavaScript trainAI function.

[1143] Step 4:

[1144] The terminal collects the input data of "Industry" and "Model Type".

[1145] Step 5:

[1146] The device uses an emotion engine to detect the user's emotional state in real time, for example by analyzing emotions from the user's facial expressions and tone of voice.

[1147] Step 6:

[1148] The device converts the collected data and emotional state into a JSON format, for example, as follows:

[1149] json

[1150] {

[1151] "industry": "Medical",

[1152] "domain_data": [], / / Include medical related data here

[1153] "model_type": "translator",

[1154] "emotional_state": "neutral" / / or "stressed", etc.

[1155] }

[1156] Step 7:

[1157] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[1158] Step 8:

[1159] The server receives the POST request and parses the JSON data.

[1160] Step 9:

[1161] The server extracts the industry, domain_data, model_type, and emotional_state fields from the parsed data.

[1162] Step 10:

[1163] The server calls the train_model function to start training the AI ​​model.

[1164] Step 11:

[1165] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[1166] Step 12:

[1167] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[1168] Step 13:

[1169] The server adjusts the training parameters based on the user's emotional state (emotional_state). For example, if the user is feeling stressed, the difficulty of the training will be reduced.

[1170] Step 14:

[1171] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[1172] Step 15:

[1173] After the server completes training, it saves the model to storage.

[1174] Step 16:

[1175] The server assigns a unique identification (model ID) to the saved model.

[1176] Step 17:

[1177] The server generates a JSON response containing a training success message and the model ID.

[1178] json

[1179] {

[1180] "message": "Training successful",

[1181] "model_id": "model12345"

[1182] }

[1183] Step 18:

[1184] The server generates a JSON response and sends it back to the device.

[1185] Step 19:

[1186] The device receives a JSON response from the server.

[1187] Step 20:

[1188] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[1189] Example 2

[1190] 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."

[1191] Conventional general-purpose AI systems have difficulty effectively learning specialized knowledge and skills specific to specific jobs or industries. Furthermore, because they do not take the user's emotional state into account, the effectiveness of training depends on the user's psychological state, resulting in reduced efficiency. Therefore, a new system that can monitor the user's emotional state in real time and optimize training is needed.

[1192] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: means for learning specialized knowledge and skills for a specific job; means for evolving a general-purpose AI into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as training data; means for monitoring the user's emotional state in real time; means for collecting data entered by the user and formatting the data in JSON format; means for transmitting the formatted data to the server; and means for analyzing the received data and starting AI training based on the data. This enables efficient training of AI specialized for a specific job, and optimal training that takes into account the user's emotional state can be achieved.

[1193] "General-purpose artificial intelligence" is a general-purpose artificial intelligence system that can be used in a wide range of areas, without being limited to specific tasks or applications.

[1194] A "job" refers to a specific task or role that a person performs in their occupation.

[1195] "Skills" is a general term for the techniques and abilities required to perform specific jobs or tasks.

[1196] An "industry expert" refers to an expert with advanced knowledge and experience in a particular industry or field.

[1197] A "specialized book" is a book that provides specialized knowledge related to a particular field or job.

[1198] A "paper" is an academic piece of writing that describes the results of a particular study or investigation.

[1199] "Records" refer to documents or data that record specific facts or events.

[1200] A "terminology glossary" is a collection of technical terms and their definitions used in a particular field or job.

[1201] An "emotion engine" refers to a system or algorithm that monitors and analyzes a user's emotional state in real time.

[1202] "Real-time monitoring" means continuously observing the user's condition in real time and immediately detecting any changes.

[1203] "Data collection" is the act or process of systematically compiling necessary information.

[1204] "JSON format" is an abbreviation for JavaScript Object Notation, and is a lightweight data exchange format for expressing data in text format.

[1205] A "server" is a computer system that receives and responds to requests from clients on a network.

[1206] A "neural network model" is a type of artificial intelligence with a structure that mimics biological nerve cells, and is an algorithm that specializes in learning and prediction for specific tasks.

[1207] "Training" is the process of providing data to an artificial intelligence or machine learning model to learn and improve its predictive capabilities and performance.

[1208] "Training parameters" refer to settings or values ​​that are adjusted during the learning process of an artificial intelligence model.

[1209] "Preprocessing" refers to the preliminary processing of raw data to convert it into a format that is easy for machine learning models to handle.

[1210] "Identification information" refers to information or an ID that uniquely identifies a specific object or data.

[1211] A "prompt" is a document or instruction used as input to a generative artificial intelligence model.

[1212] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills for a specific job, and also combines it with an emotion engine that recognizes the user's emotions. The system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[1213] Users first select the specific job or industry they want to provide expertise in and enter "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, they click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine, which analyzes the user's emotional state and dynamically adjusts training parameters.

[1214] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[1215] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[1216] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[1217] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[1218] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[1219] Take the training of medical translation AI as a concrete example. When users train a medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[1220] Example prompt sentence:

[1221] "Train a medical translator AI model using data from medical books, research papers, medical records, and medical glossaries. Adjust training parameters based on user's emotional state, ensuring lower difficulty when the user feels stressed."

[1222] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1223] Step 1: User Input

[1224] Users use the system's interface to select the specific job or industry they want to train in by entering "Medical" in the "Industry" field and "translator" in the "Model Type" field. The emotion engine monitors the user's emotional state in real time and collects that data. The input data consists of "Industry," "Model Type," and "Emotional State." For example, if the user is stressed, the "Emotional State" is recorded as "stressed."

[1225] Step 2: Data formatting and transmission by the terminal

[1226] The device collects the "Industry" and "Model Type" input by the user, as well as the "Emotional State" information obtained from the emotion engine, and formats it in JSON format. For example, the following JSON data is generated:

[1227] json

[1228] {

[1229] "Industry": "Medical",

[1230] "Model Type": "translator",

[1231] "Emotional State": "stressed"

[1232] }

[1233] This JSON data is sent as a POST request to the server's / train-ai endpoint. Data formatting and transmission are automated by the device.

[1234] Step 3: Data reception and analysis by the server

[1235] The server receives the POST request sent from the device and parses the included JSON data. The parsed results include "Industry," "Model Type," and "Emotional State." The server first parses the data and obtains the values ​​of each field. For example, "Industry" is "Medical," "Model Type" is "translator," and "Emotional State" is "stressed."

[1236] Step 4: Selecting an AI model and adjusting its parameters

[1237] The server selects an appropriate neural network model for training based on the analyzed data. In this case, because the "Model Type" is "translator," a model specialized for translation is selected. The server also references the "Emotional State" obtained from the emotion engine and adjusts parameters, such as lowering the difficulty of training if the user is feeling stressed.

[1238] Step 5: Preprocess the data and run training

[1239] The server preprocesses the domain data and converts it into a trainable format. For example, it cleans and formats data from medical papers and diagnostic reports. This preprocessed data is then used to train the AI ​​model. Specifically, the data is fed into a neural network, which learns it epoch by epoch.

[1240] Step 6: Save and notify training results

[1241] Once training is complete, the server saves the trained model with a unique identifier (model ID). For example, the trained model is assigned an ID of "model12345". The server then generates a JSON response containing a training success message and the model's identifier, and sends it to the device. An example response is as follows:

[1242] json

[1243] {

[1244] "message": "Training successful",

[1245] "modelID": "model12345"

[1246] }

[1247] Step 7: Displaying the results on your device

[1248] The device receives the response from the server and analyzes its contents. The analysis results include a success message and a model ID. The device displays this information to the user. For example, it might display "Training successful. Model ID: model12345." By checking this result, the user can confirm that the training was successful and obtain the model's identification information.

[1249] (Application example 2)

[1250] 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."

[1251] Conventional general-purpose AI training systems have had the problem of being difficult to efficiently teach specialized knowledge and skills specialized for specific jobs. Furthermore, the lack of feedback based on the user's emotional state can sometimes prevent the training from being fully effective. Furthermore, there was no system that could recommend appropriate content taking the user's emotional state into account, resulting in a less than satisfactory user experience.

[1252] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1253] In this invention, the server includes: a means for introducing a general-purpose artificial intelligence (AI) to learn specialized knowledge and skills for a specific job; a means for evolving the AI ​​into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; a means for using specialized books, papers, records, and technical glossaries as training data for the AI; a means for incorporating an emotion engine that analyzes a user's emotional state in real time and provides appropriate feedback according to that state; and a means for generating prompts and adjusting the training of the AI ​​model based on the emotional state. This enables the AI ​​to efficiently learn specialized knowledge and skills specific to a specific job, and provides feedback according to the emotional state to improve the effectiveness of training. Furthermore, the user experience can be improved by recommending content based on the user's emotional state.

[1254] "General artificial intelligence" is artificial intelligence that can handle a wide range of jobs and tasks.

[1255] "Expertise" refers to detailed information or knowledge about a particular job or field.

[1256] "Skills" refer to the techniques and abilities required to perform a particular job.

[1257] An "industry expert" is a person or group of people who have a high level of expertise in a particular job function or field.

[1258] An "emotion engine" is a software or hardware system for detecting and analyzing a user's emotional state.

[1259] "Real-time" refers to a timeframe in which processing and feedback occurs almost immediately.

[1260] A "prompt" is input data used to give instructions or ask questions to artificial intelligence.

[1261] "Training parameters" are settings or values ​​that can be adjusted during the learning process of an artificial intelligence model.

[1262] A "specialized book" is a book that contains detailed information about a specific job or field.

[1263] A "paper" is an academic document that summarizes research results and considerations on a specific topic.

[1264] A "record" is a written document that organizes past events and data.

[1265] A "terminology glossary" is a dictionary or list of technical terms related to a particular job or field.

[1266] A "content recommendation system" is a system that suggests appropriate content based on a user's specific state or preferences.

[1267] "Collaborative filtering" is a method for recommending new content based on a user's preferences and behavior.

[1268] "Content-based filtering" is a method of recommending suitable content to a user based on the characteristics of the content itself.

[1269] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[1270] User operations

[1271] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a medical translation AI, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[1272] Device behavior

[1273] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[1274] Server Operation

[1275] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[1276] Training an AI model

[1277] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[1278] How the recommendation system works

[1279] Furthermore, the server also has a content recommendation system based on the user's emotional state. For example, if the user's emotional state is estimated to be a desire to relax, it will recommend relaxing content. The recommendation algorithm uses collaborative filtering and content-based filtering.

[1280] Training results storage and notification

[1281] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[1282] Displaying results on a terminal

[1283] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[1284] Specific examples

[1285] For example, the following prompts can be input to a generative AI model:

[1286] Prompt Sentence Examples

[1287] It is assumed that the user's current emotional state is that they want to relax. Please recommend three relaxing video or music content.

[1288] This allows the system to efficiently provide relaxing content to the user and provide an optimal content viewing experience.

[1289] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1290] Step 1:

[1291] The user enters the "Industry" and "Model Type" through the interface and clicks the "Train AI" button.

[1292] Input: User-entered industry (e.g., "Medical") and model type (e.g., "translator").

[1293] Output: Data formatted in JSON format.

[1294] What happens: The user initiates a training request by entering the required information on the device screen.

[1295] Step 2:

[1296] The device collects the user's input data and the emotional state obtained from the emotion engine, and generates JSON data.

[1297] Inputs: User-entered industry, model type, and sentiment state from the sentiment engine.

[1298] Output: JSON data to send to the server.

[1299] Specific operation: The device combines the input information and the emotion data obtained from the emotion engine into a single data format (JSON).

[1300] Step 3:

[1301] The device sends the generated JSON data to the server's / train-ai endpoint as a POST request.

[1302] Input: Data in JSON format.

[1303] Output: Data is sent to the server.

[1304] What it does: Sends formatted data to the server using a POST request.

[1305] Step 4:

[1306] The server parses the received JSON data and extracts the included "Industry," "Domain Data," "Model Type," and "Emotional State."

[1307] Input: JSON data sent from the terminal.

[1308] Output: Data for each parsed field.

[1309] Specific operation: The server analyzes the data and extracts the necessary information individually.

[1310] Step 5:

[1311] The server starts training the AI ​​model based on the received data.

[1312] Input: Parsed "Industry", "Domain Data", "Model Type", "Emotional State".

[1313] Output: The trained AI model.

[1314] Specific operation: The server selects the specified neural network model and adjusts the training parameters based on the user's emotional state.

[1315] Step 6:

[1316] The server preprocesses the Domain Data and converts it into a trainable format.

[1317] Input: Raw "Domain Data".

[1318] Output: Preprocessed data.

[1319] Specific operation: Performs preprocessing such as data cleaning and normalization to prepare the data in a format that can be used by the model.

[1320] Step 7:

[1321] Save the trained AI model and generate a unique identifier (model ID).

[1322] Input: A trained AI model.

[1323] Output: Model ID.

[1324] Specific operation: The trained AI model is saved in a database and an ID is generated to identify the model.

[1325] Step 8:

[1326] The server generates a JSON response containing a training result success message and the model ID and sends it to the device.

[1327] Input: Training result, model ID.

[1328] Output: JSON response.

[1329] Specific operation: The server organizes the obtained model ID and training results, generates a response message, and sends it to the terminal.

[1330] Step 9:

[1331] The terminal receives the response from the server, analyzes the content, and displays it to the user.

[1332] Input: JSON response from the server.

[1333] Output: A success message and the model ID.

[1334] What it does: The device converts the received message into a user-friendly format and displays it on the screen, for example, "Training successful. Model ID: model12345."

[1335] 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.

[1336] 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.

[1337] 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.

[1338] [Fourth embodiment]

[1339] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1340] 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.

[1341] 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).

[1342] 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.

[1343] 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.

[1344] 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).

[1345] 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.

[1346] 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.

[1347] 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.

[1348] 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.

[1349] 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.

[1350] 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.

[1351] 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."

[1352] This invention relates to a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The operation of the system is explained below.

[1353] User operations

[1354] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a translation AI in the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, users can click the "Train AI" button to start the AI ​​training request.

[1355] Device behavior

[1356] The terminal collects the data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," and "Model Type." The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[1357] Server Operation

[1358] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," and "Model Type." The server selects a model for training and starts training the AI ​​model based on the received data.

[1359] Training an AI model

[1360] The server selects a neural network model for training based on the received "Model Type." For example, for a translation model, it configures and applies a neural network with specific input and output sizes. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[1361] Training results storage and notification

[1362] Once training is complete, the server stores the trained model along with a unique identifier that will be used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[1363] Displaying results on a terminal

[1364] When the device receives the response from the server, it analyzes the response and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message such as "Training successful. Model ID: model12345" may be displayed.

[1365] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[1366] The processing flow will be explained below.

[1367] Step 1:

[1368] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[1369] Step 2:

[1370] The user clicks the "Train AI" button.

[1371] Step 3:

[1372] The device calls the JavaScript trainAI function.

[1373] Step 4:

[1374] The terminal collects the input data of "Industry" and "Model Type".

[1375] Step 5:

[1376] The data collected by the device is converted into JSON format. For example, it is converted into the following format.

[1377] json

[1378] {

[1379] "industry": "Medical",

[1380] "domain_data": [], / / Include medical related data here

[1381] "model_type": "translator"

[1382] }

[1383] Step 6:

[1384] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[1385] Step 7:

[1386] The server receives the POST request and parses the JSON data.

[1387] Step 8:

[1388] The server extracts the industry, domain_data, and model_type fields from the parsed data.

[1389] Step 9:

[1390] The server calls the train_model function to start training the AI ​​model.

[1391] Step 10:

[1392] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[1393] Step 11:

[1394] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[1395] Step 12:

[1396] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[1397] Step 13:

[1398] After the server completes training, it saves the model to storage.

[1399] Step 14:

[1400] The server assigns a unique identification (model ID) to the saved model.

[1401] Step 15:

[1402] The server generates a JSON response containing a training success message and the model ID.

[1403] json

[1404] {

[1405] "message": "Training successful",

[1406] "model_id": "model12345"

[1407] }

[1408] Step 16:

[1409] The server generates a JSON response and sends it back to the device.

[1410] Step 17:

[1411] The device receives a JSON response from the server.

[1412] Step 18:

[1413] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[1414] Example 1

[1415] 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."

[1416] An appropriate training process is necessary for general-purpose AI to effectively learn specialized knowledge and skills for specific tasks and evolve into AI with advanced specialized knowledge. However, conventional methods have not been able to achieve efficient training because they do not effectively utilize user input data and do not fully reflect expert knowledge and guidance. Furthermore, training results are not returned to users quickly and reliably, making them less convenient to use in actual tasks.

[1417] 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.

[1418] In this invention, the server includes: a means for a user to select a specific job or industry and initiate a training request; a means for a terminal to collect user input data, format it in JSON format, and send it to the server; a means for the server to analyze the received data, select a specific model, and initiate AI training based on the data; a means for saving the AI ​​model after completion of training and generating unique identification information; and a means for providing the user with a training result success message including the identification information. This allows the general AI to effectively learn the specialized knowledge and skills required for a specific job, and makes it possible to quickly and reliably provide the trained AI model to the user.

[1419] "General-purpose artificial intelligence" is artificial intelligence that is not limited to a specific use and can handle a variety of tasks and operations.

[1420] "Specific work" refers to specific tasks or activities related to a particular job, role, or industry.

[1421] "Expertise" refers to specialized knowledge or information related to a particular job or field.

[1422] "Skills" refers to the techniques and abilities required to perform a particular task efficiently and accurately.

[1423] An "expert" is someone who has advanced knowledge and experience in a particular job or field.

[1424] "Training data" refers to the datasets used to train artificial intelligence, including specialized literature, documents, records, glossaries, etc.

[1425] "User" refers to any person or organization that uses the System to request training of the Artificial Intelligence.

[1426] A "terminal" is a device used by a user to operate the system, and includes a computer, a smartphone, etc.

[1427] The "JSON format" is a lightweight data exchange format for structuring and representing data.

[1428] A "server" refers to a computer system that receives and processes requests from users.

[1429] "Analysis" refers to the process of understanding the data received and extracting the necessary information.

[1430] A "neural network model" is a specific algorithmic model used to train artificial intelligence, which is based on neural networks.

[1431] "Preprocessing" refers to the process for converting training data into a trainable form.

[1432] "Training" refers to the process of teaching an artificial intelligence model the ability to handle a specific task.

[1433] "Identification information" refers to information that uniquely identifies a trained artificial intelligence model.

[1434] A "success message" refers to a message that informs the user that the training is complete.

[1435] The present invention is a system for training a general-purpose artificial intelligence (AI) to have specialized knowledge and skills for a specific job. This system is composed of a user, a terminal, and a server. The detailed implementation method of the system is described below.

[1436] First, the user accesses the system's operation screen and enters data to select a specific business or industry. Specifically, the user enters the industry name in the "Industry" field and the model type in the "Model Type" field. For example, if the user wants to train a translation AI for the medical field, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to start the training request.

[1437] Next, the terminal collects the data entered by the user and formats it in JSON format. For example, it generates JSON data containing information such as "Industry," "Domain Data," and "Model Type." This data is sent from the terminal to the specified endpoint of the server as a POST request. The terminal automatically formats the data and sends it to the server.

[1438] The server receives the request sent from the terminal and analyzes the included JSON data, which includes "Industry," "Domain Data," and "Model Type." Based on the analysis results, the server selects a model for training and begins training the AI ​​model based on the data. Specifically, the server selects a neural network model according to the "Model Type" and preprocesses the received "Domain Data" to convert it into a trainable format.

[1439] The server then uses the preprocessed data to train the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. Through this process, the AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[1440] Once training is complete, the server stores the trained model along with a unique identifier, which is used to identify and use the model in the future. The server then generates and sends a response to the device containing a training success message and the model's identifier.

[1441] Finally, the device receives the response from the server, analyzes it, and displays the training results to the user. The user can see a message indicating successful training and the trained model's identification information on the device screen. For example, a message like "Training successful. Model ID: model12345" can be displayed.

[1442] Take the training of medical translation AI as a concrete example. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology glossaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. Users can then use the trained AI to efficiently and accurately translate medical papers and medical certificates.

[1443] Prompt Sentence Examples

[1444] Prompt: I have created a dataset to train a translation AI specialized for the medical field. Now, please train your AI model on this dataset.

[1445] Input data: "Industry: Medical" "Domain Data: Medical books, papers, medical records, terminology" "Model Type: translator"

[1446] Output: The AI ​​model acquires specialized skills for medical translation and can provide accurate translations.

[1447] By linking user input with server processing, this system effectively specializes artificial intelligence for specific tasks, allowing users to quickly and efficiently utilize AI models with the necessary expertise and skills.

[1448] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1449] Step 1:

[1450] The user enters data

[1451] Specific behavior:

[1452] The user accesses the system's operation screen and enters data into the fields for selecting the job and industry. Specifically, enter the industry name in the "Industry" field and the model type in the "Model Type" field. For example, enter "Medical" in "Industry" and "Translator" in "Model Type." Then, click the "Train AI" button to start the training request.

[1453] Input: User-entered industry name and model type

[1454] Output: Training request started

[1455] Step 2:

[1456] The device collects and formats the data

[1457] Specific behavior:

[1458] The terminal collects the "Industry," "Domain Data," and "Model Type" data entered by the user. It formats the collected data in JSON format and prepares it for sending to the server. The generated JSON data looks like this, for example:

[1459] {

[1460] "Industry": "Medical",

[1461] "Domain Data": "Medical books, papers, medical records, terminology",

[1462] "Model Type": "translator"

[1463] }

[1464] Input: User-entered industry name, domain data, and model type

[1465] Output: JSON format data

[1466] Step 3:

[1467] The device sends data to the server

[1468] Specific behavior:

[1469] The formatted JSON data is sent as a POST request to the server's / train-ai endpoint. The device checks whether the transmission was successful.

[1470] Input: JSON format data

[1471] Output: Sending a POST request to the server

[1472] Step 4:

[1473] The server analyzes the received data

[1474] Specific behavior:

[1475] The server receives the request sent from the device and parses the included JSON data, extracting the following information from the parsed data: "Industry," "Domain Data," and "Model Type."

[1476] Input: Received JSON data

[1477] Output: Parsed "Industry", "Domain Data", and "Model Type"

[1478] Step 5:

[1479] The server selects the training model and preprocesses the data.

[1480] Specific behavior:

[1481] The server selects an appropriate neural network model based on the extracted "Model Type," and then preprocesses the acquired "Domain Data" and converts it into a trainable format.

[1482] Input: Parsed "Model Type" and "Domain Data"

[1483] Output: Data converted into a trainable format

[1484] Step 6:

[1485] The server trains the AI ​​model

[1486] Specific behavior:

[1487] Using the preprocessed data, the server trains the AI ​​model. For example, for a medical translation model, it configures and applies a neural network with specific input and output sizes. In this process, the AI ​​model acquires skills specific to medical translation.

[1488] Input: Data converted into a trainable format

[1489] Output: A trained AI model

[1490] Step 7:

[1491] The server stores and notifies the training results

[1492] Specific behavior:

[1493] Once training is complete, the server saves the model and generates a unique identifier (Model ID). The server generates a response containing a training success message and the unique identifier and sends it to the device.

[1494] Input: A trained AI model

[1495] Output: Model identification and a success message

[1496] Step 8:

[1497] The device receives the server's response and displays it to the user.

[1498] Specific behavior:

[1499] The device receives the response from the server, analyzes its contents, and displays a message indicating successful training and the model's identification information to the user. For example, a message such as "Training successful. Model ID: model12345" is displayed.

[1500] Input: Response from the server

[1501] Output: Displaying training results to the user

[1502] (Application example 1)

[1503] 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."

[1504] In modern brick-and-mortar stores, it is difficult for customers to quickly obtain detailed product information and usage instructions. Furthermore, there is a lack of a system that properly recommends related products, which prevents customers from making optimal choices when purchasing. Furthermore, inventory cannot be checked in real time, which causes inconvenience for both customers and stores.

[1505] 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.

[1506] In this invention, the server includes: means for introducing a general-purpose artificial intelligence and learning specialized knowledge and skills for a specific job; means for evolving the general-purpose artificial intelligence into an artificial intelligence suitable for the specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence; means for recommending detailed information, usage guides, and related products related to the specific job; and means for reading barcodes to obtain product information. This allows customers to obtain product information, usage instructions, and related product recommendations in real time in stores, improving the accuracy of their purchasing decisions.

[1507] "General-purpose artificial intelligence" is artificial intelligence that is not dependent on specific tasks or jobs and can be applied in a wide range of fields.

[1508] "Specialized knowledge" is detailed and advanced knowledge acquired in a particular job or industry.

[1509] "Technology" refers to the specific techniques and skills required for a particular job or industry.

[1510] An "industry expert" is a professional with extensive experience and in-depth knowledge in a particular industry.

[1511] A "specific job" is a specific job or role performed in a particular industry or field.

[1512] A "specialty book" is a publication that provides advanced knowledge or information in a specific field.

[1513] A "paper" is a document that reports the results of detailed research and analysis on a specific research subject.

[1514] A "record" is information that preserves past facts, events, and data.

[1515] A "terminology glossary" is a dictionary or list of technical terms in a particular field.

[1516] "Detailed information" means specific and comprehensive information about a product or service.

[1517] A "usage guide" is a document that explains the specific steps and methods for using a product or service.

[1518] "Related products" are other products that are highly related to a particular product.

[1519] "Recommendation" is the act of introducing something to others based on specific conditions or requirements.

[1520] A "barcode" is a series of lines and numbers used to represent the identity of a product or item.

[1521] "Product information" is detailed data about the product's characteristics, specifications, usage, price, etc.

[1522] "Stock status" is information that indicates how much of a particular product is currently in stock.

[1523] This invention provides a system that allows customers in physical stores to quickly obtain detailed product information, usage instructions, and related product recommendations. This system employs general artificial intelligence (AI) and is trained to acquire specialized knowledge and skills corresponding to specific tasks.

[1524] Hardware and Software Configuration

[1525] The system's hardware consists of a smartphone, a server, and a barcode reader, while the software uses an API client developed in Python, TensorFlow or PyTorch for training the neural network model, and Flask or Django for providing the REST API server.

[1526] Feature details

[1527] 1. Training the general-purpose AI: The server uses specialized books, papers, records, and terminology to train the general-purpose AI with specialized knowledge and skills for specific tasks. This data is sent to the server in JSON format and used to train the AI ​​model.

[1528] 2. Product information acquisition: Using a smartphone application, the customer scans the product barcode, which retrieves the corresponding product information from the server and displays it on the smartphone screen.

[1529] 3. Providing usage guide: Based on the product ID, the server provides instructions on how to use and assemble the corresponding product. This information is also displayed through the smartphone application.

[1530] 4. Related product recommendation: The server recommends other related products based on the specific product information, and this recommendation information is also displayed on the smartphone screen.

[1531] 5. Check stock availability: The server can check the stock availability of a particular product in real time and provide that information on the smartphone screen.

[1532] Processing flow

[1533] User operation: The user operates the smartphone app and scans the barcode.

[1534] Terminal operation: Collects barcode information and sends it to the server in JSON format.

[1535] Server Action: Parse the product information based on the received barcode information and generate and return a response containing detailed information, usage guides, related product recommendations, and stock availability.

[1536] Displaying the results: The smartphone app receives the response from the server and displays it to the user.

[1537] Specific examples

[1538] For example, if a user wants to get information about the latest running shoes, they can scan the barcode with their smartphone and instantly see detailed product information, instructions for use, recommendations for related running gear, and the product's availability.

[1539] Prompt Sentence Examples

[1540] "I'm looking for the latest running shoes. I want something durable and suitable for long distance running. What are the best options?"

[1541] In this way, an intelligent system that provides advanced support for the shopping experience in physical stores will be realized.

[1542] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1543] Step 1:

[1544] A user launches a smartphone application and scans a barcode. The barcode reader is used to obtain the product's barcode information, which is then entered into the application. This barcode information is a series of numbers and letters that identify a specific product.

[1545] Step 2:

[1546] The device (smartphone) formats the acquired barcode information into JSON format. This data includes the barcode information and the user's device identification information. An HTTP POST request is generated to send the formatted JSON data to the server.

[1547] Step 3:

[1548] The server receives the JSON data sent from the device. The received data includes the product barcode information. The server analyzes this information and retrieves the corresponding product details from the database.

[1549] Step 4:

[1550] The server generates usage guides and related product data based on the acquired product details. Product usage guides are data that includes procedures and operating methods, and related product data is information on other recommended products. These data are processed on the server and converted into JSON format.

[1551] Step 5:

[1552] The server queries the inventory management system to check the real-time availability of a product. The query returns the product's stock quantity and its status. The server also converts this stock information into JSON format.

[1553] Step 6:

[1554] The server compiles the generated product details, usage guide, related product information, and stock status into a single JSON response and sends this response to the device as an HTTP response. The response contains all the necessary product information.

[1555] Step 7:

[1556] The device parses the JSON response received from the server and displays it on the screen. Users can then use the smartphone application to check product details, usage guides, related product recommendations, and stock availability, allowing them to make product selections.

[1557] These steps enable users to enjoy an efficient and high-quality shopping experience in physical stores. For example, when selecting running shoes, users can obtain optimal product information by entering a prompt such as, "I'm looking for the latest running shoes. I want something durable and suitable for long-distance running. Which product would be best?"

[1558] 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.

[1559] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[1560] User operations

[1561] Users first select the specific job or industry they want to give expertise in, entering "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, users click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[1562] Device behavior

[1563] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[1564] Server Operation

[1565] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[1566] Training an AI model

[1567] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[1568] Training results storage and notification

[1569] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[1570] Displaying results on a terminal

[1571] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[1572] As a concrete example, let's take the training of medical translation AI. When users train medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires skills specialized for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis certificate, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[1573] The processing flow will be explained below.

[1574] Step 1:

[1575] The user enters "Medical" in the "Industry" field and "translator" in the "Model Type" field.

[1576] Step 2:

[1577] The user clicks the "Train AI" button.

[1578] Step 3:

[1579] The device calls the JavaScript trainAI function.

[1580] Step 4:

[1581] The terminal collects the input data of "Industry" and "Model Type".

[1582] Step 5:

[1583] The device uses an emotion engine to detect the user's emotional state in real time, for example by analyzing emotions from the user's facial expressions and tone of voice.

[1584] Step 6:

[1585] The device converts the collected data and emotional state into a JSON format, for example, as follows:

[1586] json

[1587] {

[1588] "industry": "Medical",

[1589] "domain_data": [], / / Include medical related data here

[1590] "model_type": "translator",

[1591] "emotional_state": "neutral" / / or "stressed", etc.

[1592] }

[1593] Step 7:

[1594] The device sends a POST request to the server's / train-ai endpoint, including the generated JSON data.

[1595] Step 8:

[1596] The server receives the POST request and parses the JSON data.

[1597] Step 9:

[1598] The server extracts the industry, domain_data, model_type, and emotional_state fields from the parsed data.

[1599] Step 10:

[1600] The server calls the train_model function to start training the AI ​​model.

[1601] Step 11:

[1602] The server selects a specific training model based on the "Model Type", e.g., if model_type is "translator", it will get the translation model.

[1603] Step 12:

[1604] The server preprocesses the domain_data and converts it into a format suitable for training. This involves preprocessing such as data normalization and tokenization.

[1605] Step 13:

[1606] The server adjusts the training parameters based on the user's emotional state (emotional_state). For example, if the user is feeling stressed, the difficulty of the training will be reduced.

[1607] Step 14:

[1608] The server uses the preprocessed data to train the selected AI model, which is done by optimizing the parameters of the neural network.

[1609] Step 15:

[1610] After the server completes training, it saves the model to storage.

[1611] Step 16:

[1612] The server assigns a unique identification (model ID) to the saved model.

[1613] Step 17:

[1614] The server generates a JSON response containing a training success message and the model ID.

[1615] json

[1616] {

[1617] "message": "Training successful",

[1618] "model_id": "model12345"

[1619] }

[1620] Step 18:

[1621] The server generates a JSON response and sends it back to the device.

[1622] Step 19:

[1623] The device receives a JSON response from the server.

[1624] Step 20:

[1625] The device parses the JSON response and displays a success message and the model ID to the user, for example, "Training successful. Model ID: model12345".

[1626] Example 2

[1627] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1628] Conventional general-purpose AI systems have difficulty effectively learning specialized knowledge and skills specific to specific jobs or industries. Furthermore, because they do not take the user's emotional state into account, the effectiveness of training depends on the user's psychological state, resulting in reduced efficiency. Therefore, a new system that can monitor the user's emotional state in real time and optimize training is needed.

[1629] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes: means for learning specialized knowledge and skills for a specific job; means for evolving a general-purpose AI into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; means for using specialized books, papers, records, and technical glossaries as training data; means for monitoring the user's emotional state in real time; means for collecting data entered by the user and formatting the data in JSON format; means for transmitting the formatted data to the server; and means for analyzing the received data and starting AI training based on the data. This enables efficient training of AI specialized for a specific job, and optimal training that takes into account the user's emotional state can be achieved.

[1630] "General-purpose artificial intelligence" is a general-purpose artificial intelligence system that can be used in a wide range of areas, without being limited to specific tasks or applications.

[1631] A "job" refers to a specific task or role that a person performs in their occupation.

[1632] "Skills" is a general term for the techniques and abilities required to perform specific jobs or tasks.

[1633] An "industry expert" refers to an expert with advanced knowledge and experience in a particular industry or field.

[1634] A "specialized book" is a book that provides specialized knowledge related to a particular field or job.

[1635] A "paper" is an academic piece of writing that describes the results of a particular study or investigation.

[1636] "Records" refer to documents or data that record specific facts or events.

[1637] A "terminology glossary" is a collection of technical terms and their definitions used in a particular field or job.

[1638] An "emotion engine" refers to a system or algorithm that monitors and analyzes a user's emotional state in real time.

[1639] "Real-time monitoring" means continuously observing the user's condition in real time and immediately detecting any changes.

[1640] "Data collection" is the act or process of systematically compiling necessary information.

[1641] "JSON format" is an abbreviation for JavaScript Object Notation, and is a lightweight data exchange format for expressing data in text format.

[1642] A "server" is a computer system that receives and responds to requests from clients on a network.

[1643] A "neural network model" is a type of artificial intelligence with a structure that mimics biological nerve cells, and is an algorithm that specializes in learning and prediction for specific tasks.

[1644] "Training" is the process of providing data to an artificial intelligence or machine learning model to learn and improve its predictive capabilities and performance.

[1645] "Training parameters" refer to settings or values ​​that are adjusted during the learning process of an artificial intelligence model.

[1646] "Preprocessing" refers to the preliminary processing of raw data to convert it into a format that is easy for machine learning models to handle.

[1647] "Identification information" refers to information or an ID that uniquely identifies a specific object or data.

[1648] A "prompt" is a document or instruction used as input to a generative artificial intelligence model.

[1649] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills for a specific job, and also combines it with an emotion engine that recognizes the user's emotions. The system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[1650] Users first select the specific job or industry they want to provide expertise in and enter "Medical" in the "Industry" field and "Translator" in the "Model Type" field. Then, they click the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine, which analyzes the user's emotional state and dynamically adjusts training parameters.

[1651] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[1652] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[1653] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[1654] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[1655] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[1656] Take the training of medical translation AI as a concrete example. When users train a medical translation AI, they use training data such as medical books, papers, medical records, and terminology dictionaries. Based on this training data, the AI ​​model acquires specialized skills for medical translation and is able to provide accurate translations. For example, when a user is translating a medical paper or diagnosis, the system can sense the user's stress level and provide appropriate feedback, allowing the translation work to proceed efficiently and accurately.

[1657] Example prompt sentence:

[1658] "Train a medical translator AI model using data from medical books, research papers, medical records, and medical glossaries. Adjust training parameters based on user's emotional state, ensuring lower difficulty when the user feels stressed."

[1659] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1660] Step 1: User Input

[1661] Users use the system's interface to select the specific job or industry they want to train in by entering "Medical" in the "Industry" field and "translator" in the "Model Type" field. The emotion engine monitors the user's emotional state in real time and collects that data. The input data consists of "Industry," "Model Type," and "Emotional State." For example, if the user is stressed, the "Emotional State" is recorded as "stressed."

[1662] Step 2: Data formatting and transmission by the terminal

[1663] The device collects the "Industry" and "Model Type" input by the user, as well as the "Emotional State" information obtained from the emotion engine, and formats it in JSON format. For example, the following JSON data is generated:

[1664] json

[1665] {

[1666] "Industry": "Medical",

[1667] "Model Type": "translator",

[1668] "Emotional State": "stressed"

[1669] }

[1670] This JSON data is sent as a POST request to the server's / train-ai endpoint. Data formatting and transmission are automated by the device.

[1671] Step 3: Data reception and analysis by the server

[1672] The server receives the POST request sent from the device and parses the included JSON data. The parsed results include "Industry," "Model Type," and "Emotional State." The server first parses the data and obtains the values ​​of each field. For example, "Industry" is "Medical," "Model Type" is "translator," and "Emotional State" is "stressed."

[1673] Step 4: Selecting an AI model and adjusting its parameters

[1674] The server selects an appropriate neural network model for training based on the analyzed data. In this case, because the "Model Type" is "translator," a model specialized for translation is selected. The server also references the "Emotional State" obtained from the emotion engine and adjusts parameters, such as lowering the difficulty of training if the user is feeling stressed.

[1675] Step 5: Preprocess the data and run training

[1676] The server preprocesses the domain data and converts it into a trainable format. For example, it cleans and formats data from medical papers and diagnostic reports. This preprocessed data is then used to train the AI ​​model. Specifically, the data is fed into a neural network, which learns it epoch by epoch.

[1677] Step 6: Save and notify training results

[1678] Once training is complete, the server saves the trained model with a unique identifier (model ID). For example, the trained model is assigned an ID of "model12345". The server then generates a JSON response containing a training success message and the model's identifier, and sends it to the device. An example response is as follows:

[1679] json

[1680] {

[1681] "message": "Training successful",

[1682] "modelID": "model12345"

[1683] }

[1684] Step 7: Displaying the results on your device

[1685] The device receives the response from the server and analyzes its contents. The analysis results include a success message and a model ID. The device displays this information to the user. For example, it might display "Training successful. Model ID: model12345." By checking this result, the user can confirm that the training was successful and obtain the model's identification information.

[1686] (Application example 2)

[1687] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1688] Conventional general-purpose AI training systems have had the problem of being difficult to efficiently teach specialized knowledge and skills specialized for specific jobs. Furthermore, the lack of feedback based on the user's emotional state can sometimes prevent the training from being fully effective. Furthermore, there was no system that could recommend appropriate content taking the user's emotional state into account, resulting in a less than satisfactory user experience.

[1689] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1690] In this invention, the server includes: a means for introducing a general-purpose artificial intelligence (AI) to learn specialized knowledge and skills for a specific job; a means for evolving the AI ​​into an AI suitable for a specific job based on knowledge and guidance provided by industry experts; a means for using specialized books, papers, records, and technical glossaries as training data for the AI; a means for incorporating an emotion engine that analyzes a user's emotional state in real time and provides appropriate feedback according to that state; and a means for generating prompts and adjusting the training of the AI ​​model based on the emotional state. This enables the AI ​​to efficiently learn specialized knowledge and skills specific to a specific job, and provides feedback according to the emotional state to improve the effectiveness of training. Furthermore, the user experience can be improved by recommending content based on the user's emotional state.

[1691] "General artificial intelligence" is artificial intelligence that can handle a wide range of jobs and tasks.

[1692] "Expertise" refers to detailed information or knowledge about a particular job or field.

[1693] "Skills" refer to the techniques and abilities required to perform a particular job.

[1694] An "industry expert" is a person or group of people who have a high level of expertise in a particular job function or field.

[1695] An "emotion engine" is a software or hardware system for detecting and analyzing a user's emotional state.

[1696] "Real-time" refers to a timeframe in which processing and feedback occurs almost immediately.

[1697] A "prompt" is input data used to give instructions or ask questions to artificial intelligence.

[1698] "Training parameters" are settings or values ​​that can be adjusted during the learning process of an artificial intelligence model.

[1699] A "specialized book" is a book that contains detailed information about a specific job or field.

[1700] A "paper" is an academic document that summarizes research results and considerations on a specific topic.

[1701] A "record" is a written document that organizes past events and data.

[1702] A "terminology glossary" is a dictionary or list of technical terms related to a particular job or field.

[1703] A "content recommendation system" is a system that suggests appropriate content based on a user's specific state or preferences.

[1704] "Collaborative filtering" is a method for recommending new content based on a user's preferences and behavior.

[1705] "Content-based filtering" is a method of recommending suitable content to a user based on the characteristics of the content itself.

[1706] This invention is a system that trains a general-purpose artificial intelligence to learn specialized knowledge and skills corresponding to a specific job, and also combines it with an emotion engine that recognizes the user's emotions. This system consists of a user, a terminal, and a server, and adjusts the training content based on the user's emotional state, improving the training effectiveness of the AI ​​model.

[1707] User operations

[1708] Users first select the specific job or industry they want to give expertise in. For example, if a user wants to train a medical translation AI, they would enter "Medical" in the "Industry" field and "translator" in the "Model Type" field. Then, the user clicks the "Train AI" button to initiate the AI ​​training request. The user's emotional state is also monitored in real time by the emotion engine.

[1709] Device behavior

[1710] The terminal collects data entered by the user and formats it in JSON format. Specifically, it generates JSON data containing information on "Industry," "Domain Data," "Model Type," and "Emotional State" obtained from the emotion engine. The generated data is sent as a POST request to the server's / train-ai endpoint. The terminal automatically formats the data to be sent and sends it to the server.

[1711] Server Operation

[1712] The server receives the request sent from the device and parses the included JSON data, which includes "Industry," "Domain Data," "Model Type," and "Emotional State." The server selects a model for training and starts training the AI ​​model based on the received data.

[1713] Training an AI model

[1714] The server selects a neural network model for training based on the "Model Type." It also references the "Emotional State" obtained from the emotion engine and adjusts the training parameters according to the user's emotional state. For example, if the user is feeling stressed, it may adjust the training difficulty to lower it. Next, the server preprocesses the "Domain Data" and converts it into a trainable format. This preprocessed data is used to train the AI ​​model. As a result of the training, the model acquires the ability to perform the specified task (in this case, translating medical documents).

[1715] How the recommendation system works

[1716] Furthermore, the server also has a content recommendation system based on the user's emotional state. For example, if the user's emotional state is estimated to be a desire to relax, it will recommend relaxing content. The recommendation algorithm uses collaborative filtering and content-based filtering.

[1717] Training results storage and notification

[1718] Once training is complete, the server stores the trained model along with a unique identifier (model ID) that will be used to identify and use the model in the future. The server then generates a JSON response containing a training success message and the model's identifier and sends it to the device.

[1719] Displaying results on a terminal

[1720] When the device receives the response from the server, it analyzes the response and displays a success message and the model ID to the user. The user can see the training success message and the trained model's identification information on the device screen. For example, the message might look like this: "Training successful. Model ID: model12345."

[1721] Specific examples

[1722] For example, the following prompts can be input to a generative AI model:

[1723] Prompt Sentence Examples

[1724] It is assumed that the user's current emotional state is that they want to relax. Please recommend three relaxing video or music content.

[1725] This allows the system to efficiently provide relaxing content to the user and provide an optimal content viewing experience.

[1726] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1727] Step 1:

[1728] The user enters the "Industry" and "Model Type" through the interface and clicks the "Train AI" button.

[1729] Input: User-entered industry (e.g., "Medical") and model type (e.g., "translator").

[1730] Output: Data formatted in JSON format.

[1731] What happens: The user initiates a training request by entering the required information on the device screen.

[1732] Step 2:

[1733] The device collects the user's input data and the emotional state obtained from the emotion engine, and generates JSON data.

[1734] Inputs: User-entered industry, model type, and sentiment state from the sentiment engine.

[1735] Output: JSON data to send to the server.

[1736] Specific operation: The device combines the input information and the emotion data obtained from the emotion engine into a single data format (JSON).

[1737] Step 3:

[1738] The device sends the generated JSON data to the server's / train-ai endpoint as a POST request.

[1739] Input: Data in JSON format.

[1740] Output: Data is sent to the server.

[1741] What it does: Sends formatted data to the server using a POST request.

[1742] Step 4:

[1743] The server parses the received JSON data and extracts the included "Industry," "Domain Data," "Model Type," and "Emotional State."

[1744] Input: JSON data sent from the terminal.

[1745] Output: Data for each parsed field.

[1746] Specific operation: The server analyzes the data and extracts the necessary information individually.

[1747] Step 5:

[1748] The server starts training the AI ​​model based on the received data.

[1749] Input: Parsed "Industry", "Domain Data", "Model Type", "Emotional State".

[1750] Output: The trained AI model.

[1751] Specific operation: The server selects the specified neural network model and adjusts the training parameters based on the user's emotional state.

[1752] Step 6:

[1753] The server preprocesses the Domain Data and converts it into a trainable format.

[1754] Input: Raw "Domain Data".

[1755] Output: Preprocessed data.

[1756] Specific operation: Performs preprocessing such as data cleaning and normalization to prepare the data in a format that can be used by the model.

[1757] Step 7:

[1758] Save the trained AI model and generate a unique identifier (model ID).

[1759] Input: A trained AI model.

[1760] Output: Model ID.

[1761] Specific operation: The trained AI model is saved in a database and an ID is generated to identify the model.

[1762] Step 8:

[1763] The server generates a JSON response containing a training result success message and the model ID and sends it to the device.

[1764] Input: Training result, model ID.

[1765] Output: JSON response.

[1766] Specific operation: The server organizes the obtained model ID and training results, generates a response message, and sends it to the terminal.

[1767] Step 9:

[1768] The terminal receives the response from the server, analyzes the content, and displays it to the user.

[1769] Input: JSON response from the server.

[1770] Output: A success message and the model ID.

[1771] What it does: The device converts the received message into a user-friendly format and displays it on the screen, for example, "Training successful. Model ID: model12345."

[1772] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1773] 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.

[1774] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1775] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1776] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1777] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1778] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1779] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1780] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1781] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1782] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1783] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1784] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1785] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1786] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1787] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1788] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1789] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1790] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1791] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1792] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1793] The following is further disclosed regarding the above embodiment.

[1794] (Claim 1)

[1795] Introducing general artificial intelligence to learn specialized knowledge and skills for specific jobs;

[1796] A means for evolving the general AI into a job-specific AI based on knowledge and guidance provided by industry experts;

[1797] A means for utilizing specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence;

[1798] A system including:

[1799] (Claim 2)

[1800] A means for selecting a specialized artificial intelligence model for a specific job and training the model;

[1801] means for storing the trained model and generating an identification of the model;

[1802] 10. The system of claim 1, further comprising: means for providing said identification information to a user.

[1803] (Claim 3)

[1804] means for collecting data entered by a user and transmitting the data to a server;

[1805] means for analyzing the data received by the server and initiating training of the artificial intelligence based on the data;

[1806] 10. The system of claim 1, further comprising means for returning the results of the training to the user.

[1807] "Example 1"

[1808] (Claim 1)

[1809] Introducing general artificial intelligence to learn specialized knowledge and skills for specific tasks;

[1810] A means for evolving the general AI into an AI suitable for a specific task based on knowledge and guidance provided by experts;

[1811] A means for utilizing specialized literature, materials, records, and glossaries as learning data for the general-purpose artificial intelligence;

[1812] a means for a user to select a particular job or industry and initiate a training request;

[1813] A means for the terminal to collect user input data, format it in JSON format, and send it to the server;

[1814] means for the server to analyze the received data, select a specific model, and start training the artificial intelligence based on the data;

[1815] a means for storing the trained artificial intelligence model and generating a unique identifier;

[1816] means for providing a training result success message to the user, the message including the identification information;

[1817] A system including:

[1818] (Claim 2)

[1819] means for preprocessing the artificial intelligence model and converting it into a trainable form;

[1820] means for using the data to train an artificial intelligence;

[1821] means for storing the artificial intelligence model and generating and providing a unique identification to the user after training is complete;

[1822] means for storing said identification information for future use;

[1823] The system of claim 1 further comprising:

[1824] (Claim 3)

[1825] means for collecting data entered by a user and transmitting the data to a server;

[1826] means for analyzing the data received by the server and initiating training of the artificial intelligence based on the data;

[1827] 10. The system of claim 1, further comprising means for returning the results of the training to the user.

[1828] "Application Example 1"

[1829] (Claim 1)

[1830] Introducing general artificial intelligence to learn specialized knowledge and skills for specific jobs;

[1831] A means for evolving the general AI into a job-specific AI based on knowledge and guidance provided by industry experts;

[1832] A means for utilizing specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence;

[1833] A means of providing detailed information, how-to guides and related product recommendations for specific tasks;

[1834] A means for reading the barcode and obtaining product information;

[1835] A system including:

[1836] (Claim 2)

[1837] A means for selecting a specialized artificial intelligence model for a specific job and training the model;

[1838] means for storing the trained model and generating an identification of the model;

[1839] means for providing said identification information to a user;

[1840] 10. The system of claim 1, further comprising means for checking availability of a particular product.

[1841] (Claim 3)

[1842] means for collecting data entered by a user and transmitting the data to a server;

[1843] means for analyzing the data received by the server and initiating training of the artificial intelligence based on the data;

[1844] means for returning the results of the training to the user;

[1845] 10. The system of claim 1, further comprising means for obtaining product information by barcode scanning and recommending related products.

[1846] "Example 2: Combining Emotion Engines"

[1847] (Claim 1)

[1848] Introducing general artificial intelligence to learn specialized knowledge and skills for specific jobs;

[1849] A means for evolving the general AI into a job-specific AI based on knowledge and guidance provided by industry experts;

[1850] A means for utilizing specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence;

[1851] means for the device to monitor the user's emotional state in real time;

[1852] means for adjusting training parameters based on the user's emotional state;

[1853] A means for collecting data entered by a user and formatting the data in JSON format;

[1854] means for transmitting the formatted data to a server;

[1855] A system including:

[1856] (Claim 2)

[1857] A means for selecting a specialized artificial intelligence model for a specific job and training the model;

[1858] means for storing the trained model and generating an identification of the model;

[1859] means for providing said identification information to a user;

[1860] means for returning the training results of the model to the user along with a success message;

[1861] 10. The system of claim 1, further comprising: means for displaying a training success message and an identification of the model to the user.

[1862] (Claim 3)

[1863] means for analyzing the data received by the server and initiating training of the artificial intelligence based on the data;

[1864] The analyzed data includes "job role", "domain data", "model type" and "emotional state";

[1865] means for storing an identification of the trained model and making the identification available;

[1866] means for analyzing the content of a response received by the terminal from the server and displaying the training results to the user;

[1867] The system of claim 1 further comprising:

[1868] "Application example 2 when combining emotion engines"

[1869] (Claim 1)

[1870] Introducing general artificial intelligence to learn specialized knowledge and skills for specific jobs;

[1871] A means for evolving the general AI into a job-specific AI based on knowledge and guidance provided by industry experts;

[1872] A means for utilizing specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence;

[1873] A means for incorporating an emotion engine that analyzes the user's emotional state in real time and provides appropriate feedback according to that state;

[1874] A means for generating prompts to tailor the training of the AI ​​model based on emotional state; and

[1875] A system including:

[1876] (Claim 2)

[1877] A means for selecting a specialized artificial intelligence model for a specific job and training the model;

[1878] means for storing the trained model and generating an identification of the model;

[1879] means for providing said identification information to a user;

[1880] 10. The system of claim 1, further comprising means for recommending content based on an emotional state of the user.

[1881] (Claim 3)

[1882] means for collecting data entered by a user and transmitting the data to a server;

[1883] means for analyzing the data received by the server and initiating training of the artificial intelligence based on the data;

[1884] means for returning the results of the training to the user;

[1885] 10. The system of claim 1, further comprising means for providing recommended content in response to an emotional state obtained from the emotional engine. [Explanation of symbols]

[1886] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. Introducing general artificial intelligence to learn specialized knowledge and skills for specific jobs; A means for evolving the general AI into a job-specific AI based on knowledge and guidance provided by industry experts; A means for utilizing specialized books, papers, records, and technical glossaries as learning data for the general-purpose artificial intelligence; A system including:

2. A means for selecting a specialized artificial intelligence model for a specific job and training the model; means for storing the trained model and generating an identification of the model; 2. The system of claim 1, further comprising means for providing said identification information to a user.

3. means for collecting data entered by a user and transmitting the data to a server; means for analyzing the data received by the server and initiating training of the artificial intelligence based on the data; 10. The system of claim 1, further comprising means for returning the results of the training to the user.

Images