system

Abstract

Provide a system. [Solution] A means of collecting data related to a particular industry; means for training an interactive artificial intelligence model using the collected data; A means to customize your company's business processes and terminology, means for receiving input data from a user and storing an interaction log; A means of analyzing the accumulated dialogue logs and retraining the model; A means to develop and update new industry-specific dialogue packages; means for providing an updated interaction package to a user; A system including:

Description

[Technical Field]

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

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

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

[0004] ---

[0005] Conventional AI systems for corporate industries were general, making it difficult to provide specialized services tailored to specific industries. They also lacked the ability to customize to suit business processes or the ability to accumulate and analyze user data to improve model accuracy. As a result, they were unable to sufficiently improve business efficiency, and were particularly unpractical in highly specialized industries. To solve this problem, we aim to build an interactive artificial intelligence system optimized for specific industries and provide a system capable of providing specialized and efficient business support. [Means for solving the problem]

[0006] The present invention provides a system including means for collecting data related to a specific industry, means for training an interactive AI model using the collected data, means for customizing a company's business processes and terminology, means for receiving input data from a user and accumulating dialogue logs, means for analyzing the accumulated dialogue logs and retraining the model, means for developing and updating a new industry-specific dialogue package, and means for providing the updated dialogue package to the user. This system realizes a highly accurate interactive AI tailored to the expertise of a specific industry, thereby significantly improving business efficiency.

[0007] ---

[0008] "Industry-related data" refers to information such as terminology, business processes, best practices, and training materials related to a specific industry.

[0009] An "interactive artificial intelligence model" refers to an AI model that uses natural language processing technology to engage in dialogue with users and generate responses optimized for specific industries.

[0010] "Customization methods" refers to methods and tools for adapting a conversational AI model based on a specific company's business processes and terminology.

[0011] "User input data" refers to information such as questions, instructions, and feedback that a user inputs into an interactive AI system.

[0012] A "dialogue log" refers to the history of a series of conversations between a user and an interactive AI system, and accumulating this information is useful for improving and analyzing models.

[0013] "Means of analyzing data" refers to methods and technologies for analyzing accumulated dialogue logs and improving the accuracy of the model based on the insights gained from them.

[0014] "New industry-specific dialogue packages" refer to software packages that include new AI dialogue models and functions developed specifically for specific industries.

[0015] "Means of delivery" refers to the methods and processes for distributing newly developed interaction packages to companies and updating their systems to make them available. [Brief explanation of the drawings]

[0016] [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

[0017] 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.

[0018] First, the terms used in the following description will be explained.

[0019] 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).

[0020] 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.

[0021] 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.

[0022] 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.

[0023] 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."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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."

[0037] ---

[0038] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[0039] server

[0040] The server provides the following means:

[0041] 1. Data collection method: Collect data related to a specific industry from external databases and specialized books, thereby understanding industry-specific terminology and business processes.

[0042] 2. Model training method: Based on the collected data, an interactive artificial intelligence model is trained, which is built to respond appropriately to industry-specific questions and requests.

[0043] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[0044] 4. Package development method: Based on the analysis results, we develop new industry-specific dialogue packages, thereby providing customized solutions for each company.

[0045] Terminal

[0046] The terminal provides the means to:

[0047] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[0048] 2. Data reception and transmission means: Sends input data from the user to the server in real time, and presents response data from the server to the user.

[0049] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[0050] User

[0051] The user provides the means to:

[0052] 1. Input: Using conversational AI in everyday tasks, you input questions and commands into the system, such as specific requests like "What's on the agenda for today's meeting?" or "Tell me about customer A's past purchase history."

[0053] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[0054] Example: AI system for medical institutions

[0055] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[0056] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[0057] The terminal works in conjunction with the hospital's systems and provides an interface for setting commonly used terms and medical procedures for doctors and nurses.

[0058] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, and the server generates an appropriate response and sends it to the terminal.

[0059] The server collects and analyzes the dialogue logs between users and the AI. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[0060] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[0061] In this way, the present invention provides an interactive artificial intelligence system optimized for a specific industry, which can greatly improve the business efficiency of an enterprise.

[0062] The processing flow will be explained below.

[0063] ---

[0064] Step 1:

[0065] The server collects data related to a particular industry from external databases and specialized books.

[0066] Step 2:

[0067] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[0068] Step 3:

[0069] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[0070] Step 4:

[0071] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[0072] Step 5:

[0073] Users input their company's specific business processes and terminology through the interface.

[0074] Step 6:

[0075] The server collects input data from users and reflects it in the AI ​​model.

[0076] Step 7:

[0077] Users use conversational artificial intelligence in their daily work, for example, by typing, "What's on the agenda for today's meetings?"

[0078] Step 8:

[0079] The terminal transmits the user's input to the server in real time.

[0080] Step 9:

[0081] The server generates a response based on the user's input and sends the response to the terminal.

[0082] Step 10:

[0083] The terminal presents the generated response to the user.

[0084] Step 11:

[0085] The server securely stores the dialogue logs between the user and the AI.

[0086] Step 12:

[0087] The server periodically retrieves the accumulated dialogue logs and performs data preprocessing.

[0088] Step 13:

[0089] The server analyzes the pre-processed data and identifies areas for improvement in the model.

[0090] Step 14:

[0091] The server retrains the conversational AI model based on the analysis results.

[0092] Step 15:

[0093] The server develops new industry-specific dialogue packages.

[0094] Step 16:

[0095] The terminal receives new interaction packages from the server and updates the system.

[0096] Step 17:

[0097] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[0098] Step 18:

[0099] The device sends the user's input to the server, and the conversational AI generates a response and returns it to the device.

[0100] Step 19:

[0101] The terminal presents the new response to the user.

[0102] Step 20:

[0103] The server continues to store updated interaction logs to help improve the accuracy of the system.

[0104] ---

[0105] These are the specific steps in programming an interactive AI system optimized for a specific industry.

[0106] Example 1

[0107] 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."

[0108] Conventional conversational AI systems have struggled to consistently provide everything from data collection to model training and customization, dialogue log accumulation and analysis, and automatic system updates to meet the specialized requirements of specific industries. As a result, they have been unable to provide highly accurate responses tailored to each company's needs. While this is expected to improve business efficiency, it actually places a heavy burden on users and lacks convenience. In particular, there is a need for appropriate and rapid response in the process of generating real-time responses, refining models, and updating dialogue packages based on feedback.

[0109] 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.

[0110] In this invention, the server includes a means for collecting data related to a specific industry, a means for training an interactive AI model using the collected data, and a means for generating responses to users. This makes it possible to collect data suited to the specialized needs of each company and generate highly accurate responses in real time. Furthermore, by supporting customization, analysis of dialogue logs, model retraining based on feedback, and automatic updates of dialogue packages, it is possible to improve business efficiency and reduce the burden on users.

[0111] "Means for collecting data related to a specific industry" refers to methods or devices for obtaining information related to a specific industry from external databases or specialized books.

[0112] "Means for training an interactive AI model using collected data" refers to a method or device for building an interactive AI model using a machine learning algorithm based on acquired data and optimizing the model.

[0113] "Means for customizing a company's business processes and terminology" refers to methods and devices that provide interfaces and configuration functions for reflecting the unique business procedures and terminology of each company within the system.

[0114] "Means for receiving input data from a user and transmitting it to a server in real time" refers to a communication protocol or device for instantly receiving data input by a user into the system and transmitting it to a server.

[0115] A "means for generating a response to a user" is a method or device that utilizes an interactive artificial intelligence model to generate an appropriate reply based on input data from a user.

[0116] "Means for analyzing dialogue logs and retraining models" refers to a method or device that analyzes dialogue history with a user and retrains the model based on this to improve the accuracy of the dialogue-based artificial intelligence model.

[0117] "Means for developing and updating new industry-specific dialogue packages" refers to methods and devices for creating new dialogue packages tailored to specific industries based on analysis results and feedback, and for keeping the system up to date.

[0118] "Means for delivering updated dialogue packages to terminals and automatically updating the system" refers to a method or device for applying the latest dialogue packages delivered from a server to terminals and automatically updating the system.

[0119] MODE FOR CARRYING OUT THE INVENTION

[0120] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[0121] server

[0122] The server provides the following means:

[0123] 1. A means of collecting data related to a specific industry: The server obtains information related to a specific industry from external databases or specialized books. For example, in the medical industry, data is collected from PubMed and other medical databases. During this process, the data is acquired and analyzed using Python's requests library, BeautifulSoup, and Scrapy.

[0124] 2. Means for training an interactive AI model using collected data: The server uses the collected data to train an interactive AI model using Tensorflow (registered trademark) or PyTorch. Specifically, it uses natural language processing (NLP) technology to build a model that can appropriately respond to questions and requests specific to the industry.

[0125] 3. Means for generating a response to the user: The server uses an interactive artificial intelligence model to generate an appropriate response based on the input data from the user. For example, if a doctor inputs "Please tell me about patient X's allergies," the server will query the electronic medical record database and generate an appropriate response from the results.

[0126] 4. Analyzing the dialogue log and retraining the model: The server analyzes the dialogue history with the user and retrains the dialogue AI model to improve its accuracy. It uses Python's pandas and NumPy to analyze the log data and retrains the model based on the analysis results.

[0127] 5. Means for developing and updating new industry-specific dialogue packages: The server creates new dialogue packages tailored to specific industries based on analysis results and feedback, and updates the system to keep it up to date.

[0128] Terminal

[0129] The terminal provides the means to:

[0130] 1. A means to customize a company's business processes and terminology: The terminal provides an interface that reflects a company's specific business procedures and terminology within the system. For example, a customized interface can be built using a front-end framework such as React or Vue.js.

[0131] 2. Means for receiving input data from the user and transmitting it to the server in real time: The terminal uses a communication protocol (e.g., REST API or WebSocket) to transmit data input by the user to the server in real time.

[0132] 3. Means for delivering updated dialogue packages to terminals and automatically updating the system: The terminals receive the latest dialogue packages delivered from the server and automatically update the system, allowing users to always interact with the latest information.

[0133] User

[0134] The user provides the means to:

[0135] 1. Input means: A user uses an interactive AI system to input specific questions or commands into the system. For example, a specific request such as "Please tell me the allergy information of patient X" is input into a terminal.

[0136] 2. Feedback: The user provides feedback on the system's response. They evaluate whether the response is accurate or there is room for improvement, and send the feedback to the server via their device. This feedback is used to retrain the model and improve the system.

[0137] Example: AI system for medical institutions

[0138] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[0139] The server collects data on medical terminology and medical procedures from medical databases and uses TensorFlow to train an interactive AI model. For example, it can collect research paper data from PubMed via an API and use it to build a neural network model.

[0140] The device connects to the hospital's electronic medical record system and provides an interface for doctors and nurses to set commonly used terms and medical procedures. This is achieved through a web application written in React.

[0141] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, which retrieves the data from the electronic medical record, generates an appropriate response, and returns it to the terminal.

[0142] The server collects dialogue logs between users and the AI ​​and analyzes them using Python's pandas. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[0143] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[0144] Examples of prompt statements

[0145] Below are some example prompts from an AI model for healthcare:

[0146] "I'd like to see Patient X's most recent medical records. What procedures were performed?"

[0147] By inputting this prompt into a generative AI model, it is possible to extract appropriate medical information. Furthermore, it can also suggest improvements to the prompt based on the results of analyzing actual dialogue logs.

[0148] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0149] Step 1: Data Collection (Server)

[0150] Specific operation: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses Python's requests library to send an HTTP request to the PubMed API and parses the returned JSON response.

[0151] Input: Data related to a specific industry (e.g., PubMed API)

[0152] Data processing: HTTP request sending, response analysis

[0153] Output: Retrieved data (JSON format)

[0154] Step 2: Model training (server)

[0155] Specific operation: The server uses the collected data to train an interactive artificial intelligence model. For example, it uses the TensorFlow library to build a neural network model and supplies training data to train the model.

[0156] Input: Collected data (JSON format)

[0157] Data processing: Data preprocessing (cleaning, tokenization), model building, training

[0158] Output: A trained interactive AI model

[0159] Step 3: Interface Settings (Device)

[0160] Specific operation: The terminal provides an interface for inputting a company's business processes and terminology into the system. The terminal uses React to build a customized interface, allowing users to input company-specific information.

[0161] Input: Company business processes and terminology (user input)

[0162] Data processing: Interface construction

[0163] Output: Customized configuration information

[0164] Step 4: User Input (User)

[0165] Specific Actions: A user (e.g., a doctor or nurse) uses a conversational artificial intelligence system to input specific questions or commands into the system, and transmits data to the system using text input or voice recognition technology (e.g., Google® Speech-to-Text API).

[0166] Input: User questions or commands (text or voice)

[0167] Data processing: Voice recognition (when inputting voice)

[0168] Output: User input data (text format)

[0169] Step 5: Send data (terminal)

[0170] Specific operation: The terminal sends user input data to the server in real time. To do this, the terminal uses REST API or WebSocket to send data with low latency.

[0171] Input: User input data (text format)

[0172] Data processing: Data encoding, HTTP POST request or WebSocket transmission

[0173] Output: Data sent to the server

[0174] Step 6: Response Generation (Server)

[0175] Specific operation: The server generates a response using an interactive artificial intelligence model based on the received user input data. The server sends a query to the electronic medical record database and generates an appropriate response from the results.

[0176] Input: User input data (text format)

[0177] Data processing: query generation, database interrogation, response generation

[0178] Output: Generated response data (text format)

[0179] Step 7: Response presentation (terminal)

[0180] Specific operation: The terminal presents the response data returned from the server to the user. The terminal dynamically updates the web page and displays the response in a user-friendly format.

[0181] Input: Response data from the server (text format)

[0182] Data processing: Data decoding, display format conversion

[0183] Output: The response presented to the user

[0184] Step 8: Feedback (User)

[0185] Specific operation: The user provides feedback on the system's response by entering an evaluation and suggestions for improvement through the terminal and sending the feedback to the server.

[0186] Input: User feedback (text format)

[0187] Data processing: Evaluation data input

[0188] Output: Feedback sent to the server

[0189] Step 9: Log analysis and model update (server)

[0190] How it works: The server analyzes the collected interaction logs and feedback. It uses Python's pandas and NumPy to extract insights from the log data and retrain the model.

[0191] Input: Dialogue logs, feedback data

[0192] Data processing: log analysis, data preprocessing, model retraining

[0193] Output: Updated interactive artificial intelligence model

[0194] Step 10: Update packages (device)

[0195] Specific operation: The terminal receives new industry-specific dialogue packages distributed by the server and automatically updates the system, so that the latest dialogue models and functions are applied to the terminal.

[0196] Input: New dialogue package

[0197] Data processing: Package download, update application

[0198] Output: Updated system

[0199] (Application example 1)

[0200] 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."

[0201] In modern business, there is a growing demand for conversational AI systems to efficiently handle specific industries. In particular, the food delivery industry needs ways to streamline order taking, delivery tracking, and customer support. However, performing these tasks manually is time-consuming, labor-intensive, and prone to errors. Therefore, the present invention aims to provide an industry-specific conversational AI system to solve these problems.

[0202] 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.

[0203] In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive AI model using the collected data, means for customizing a company's business processes and terminology, means for receiving input data from a user and accumulating an interaction log, means for analyzing the accumulated interaction log and retraining the model, means for developing and updating a new industry-specific interaction package, means for providing the updated interaction package to the user, means for converting user voice input into text, means for analyzing user order information using the interactive AI model and providing menu options, and means for tracking delivery status for a specific order and providing the user with an estimated delivery time. This makes it possible to streamline a series of operations, from order taking to delivery tracking and customer support, even in the food delivery industry.

[0204] "Specific industry" refers to all businesses related to a specific business area or field.

[0205] "Data collection" refers to the process of obtaining information relevant to a particular industry.

[0206] "Conversational artificial intelligence model" refers to a machine learning model that is trained to respond appropriately to user input in natural language.

[0207] "A company's business processes" refers to the set of procedures and methods by which a company carries out its daily business operations.

[0208] "Jargon" refers to words and phrases specific to a particular industry or field.

[0209] "Customization" refers to the process of tailoring a system or service to meet specific needs and requirements.

[0210] "Input Data" refers to information or instructions provided by a user.

[0211] "Dialogue log" refers to a record of all conversations exchanged between a user and an interactive AI system.

[0212] "Data analysis" refers to the process of analyzing accumulated data and extracting meaningful information and patterns.

[0213] "Model retraining" refers to the process of retraining an existing machine learning model based on new data or analytical results.

[0214] "Dialogue package" refers to a software module that forms part of a dialogue-based artificial intelligence system optimized for a specific industry.

[0215] "Voice input" refers to the process of capturing user spoken words into the system.

[0216] "Menu options" refer to selectable products and services offered to a user.

[0217] "Delivery status" refers to the progress of an ordered item until it is delivered.

[0218] "Expected Delivery Time" means the estimated time it will take for an ordered item to be delivered.

[0219] The present invention provides an interactive artificial intelligence system optimized for a specific industry, streamlining order acceptance, delivery tracking, and customer support in the food delivery industry. This system is configured to operate in cooperation with each element: server, terminal, and user.

[0220] server

[0221] The server provides the following means:

[0222] 1. Data collection method: Collect information related to a specific industry from external databases to understand industry-specific terminology and business processes.

[0223] 2. Model training method: Based on the collected data, a conversational artificial intelligence model (generative AI model) is trained, which builds a model that can respond appropriately to industry-specific questions and requests.

[0224] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[0225] 4. Package development tools: Based on the analysis results, we will develop new industry-specific interactive packages to optimize order taking, delivery tracking, and customer support in the food delivery industry.

[0226] Terminal

[0227] The terminal provides the means to:

[0228] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[0229] 2. Data reception and transmission means: Data input from the user (e.g., order details or questions) is transmitted to the server in real time, and response data from the server is presented to the user.

[0230] 3. Voice input means: Provides a voice recognition function to convert voice input into text.

[0231] 4. Package update means: Receive new industry-specific dialogue packages (generative AI models) distributed from the server and update the system.

[0232] User

[0233] The user provides the means to:

[0234] 1. Input method: Food delivery orders and inquiries are input into the conversational AI system. For example, you can say, "I'd like to order one Margherita pizza and one Coke Zero" by voice.

[0235] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[0236] Hardware and software used

[0237] Hardware: Cloud servers (AWS (registered trademark), Google Cloud, etc.), smartphones (iOS, ANDROID (registered trademark), etc.)

[0238] Software: Generative AI model (facebook / blenderbot-400M-distill), framework (Flask), speech recognition API (Google Cloud Speech-to-Text), database (MongoDB)

[0239] Data processing and calculation

[0240] Speech-to-text conversion: Uses the Google Cloud Speech-to-Text API to convert speech to text, allowing you to capture user speech as text.

[0241] Dialogue model response generation: A generative AI model trained on collected data is used to generate appropriate responses to user input. For example, if a user inputs, "I'd like to order one Margherita pizza and one Coke Zero," the conversational AI system will provide corresponding menu options.

[0242] Delivery status tracking and provision: Real-time delivery status information for specific orders is tracked and provided to users. For example, if a user requests "What is the delivery status of order number 12345?", the current delivery status and estimated delivery time will be provided.

[0243] Specific examples

[0244] Order Taker: "I'd like to order one Margherita pizza and one Coke Zero."

[0245] Delivery tracking: "What's the delivery status of order number 12345?"

[0246] Customer Support: "Why can't I use my coupon code?"

[0247] Prompt Sentence Examples

[0248] Enter your order: "I'd like to order one Margherita pizza and one Coke Zero."

[0249] Delivery tracking: "What's the delivery status of order number 12345?"

[0250] Support Enquiry: "Why can't I use my coupon code?"

[0251] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0252] Step 1:

[0253] Users voice-order food delivery through a smartphone application, where the voice input is captured.

[0254] Input: User speaks "I would like to order one Margherita pizza and one Coke Zero."

[0255] Output: Audio input data

[0256] Step 2:

[0257] The device converts voice input into text using the Google Cloud Speech-to-Text API.

[0258] Input: Voice input data

[0259] Output: Text data "I would like to order one Margherita pizza and one Coke Zero."

[0260] Specific operation: Call the voice recognition API and obtain the voice data as text data.

[0261] Step 3:

[0262] The device sends text data to the server, which uses a generative AI model (facebook / blenderbot-400M-distill) to generate an appropriate response to the user's order.

[0263] Input: Text data "I would like to order one Margherita pizza and one Coke Zero."

[0264] Output: Response data containing menu options

[0265] Specific operation: Text data is input into an interactive artificial intelligence model and response text is generated.

[0266] Step 4:

[0267] The server transmits the generated response data to the terminal, which then presents it to the user.

[0268] Input: Response data "Would you like one Margherita pizza and one Coke Zero?"

[0269] Output: The response text to display to the user

[0270] Specific behavior: Display the response text on the device's user interface.

[0271] Step 5:

[0272] The user provides confirmation, for example, a voice or text reply of "Yes, please."

[0273] Input: User response "Yes, please"

[0274] Output: Input data for verification

[0275] Step 6:

[0276] The terminal performs voice recognition again and sends the result as text data to the server. The server receives this confirmation data and processes it to confirm the order.

[0277] Input: Input data for confirmation "Yes, please"

[0278] Output: Order confirmation data

[0279] Specific operation: Converts speech into text and sends the data to the server to confirm the order. The server stores the order information in a database.

[0280] Step 7:

[0281] After the order is confirmed, the server generates a tracking ID for tracking the delivery status and sends it to the terminal. The terminal displays the tracking ID to the user.

[0282] Input: Order confirmation data

[0283] Output: Tracking ID

[0284] Specific operation: A tracking ID is generated based on the order information and sent to the terminal, which then displays the tracking ID on the user interface.

[0285] Step 8:

[0286] A user queries the application to check the delivery status of an order, entering something like "What is the delivery status of order number 12345?"

[0287] Input: User inquiry: "What is the delivery status of order number 12345?"

[0288] Output: Delivery tracking request data

[0289] Step 9:

[0290] The server tracks the delivery status, obtains the status in real time, and transmits the data to the terminal, which then presents the status data to the user.

[0291] Input: Delivery tracking request data

[0292] Output: Delivery status data "Current delivery status is on its way and will arrive in 15 minutes."

[0293] Specific operation: Obtain the current status from the delivery tracking system and send the information to the terminal.

[0294] Step 10:

[0295] The order is completed when the user receives it. The user provides feedback to the system, for example, by typing "The item arrived safely."

[0296] Input: User feedback "The item arrived safely."

[0297] Output: Feedback data

[0298] Specific operation: Feedback data is sent to the server and used to improve the system.

[0299] 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.

[0300] ---

[0301] This invention is a system that provides more sophisticated responses by combining a conversational AI system optimized for a specific industry with an emotion engine that recognizes the user's emotions. This system is configured so that the server, terminal, and user elements work together and the emotion engine is used to realize responses based on the user's emotional state.

[0302] server

[0303] The server provides the following means:

[0304] 1. Data collection methods: Data related to a specific industry are collected from external databases and specialized books.

[0305] 2. Model training method: Train an interactive artificial intelligence model based on the collected data.

[0306] 3. Emotion engine: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response.

[0307] 4. Data analysis: Analyze the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights.

[0308] 5. Package development method: Based on the analysis results, new industry-specific dialogue packages will be developed.

[0309] Terminal

[0310] The terminal provides the means to:

[0311] 1. Customization methods: Provide an interface for inputting a company's business processes and terminology into the system.

[0312] 2. Data reception and transmission means: Input data and emotion data from the user are transmitted to the server in real time, and response data from the server is presented to the user.

[0313] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[0314] User

[0315] The user provides the means to:

[0316] 1. Input means: Use conversational artificial intelligence in your daily work by inputting questions and commands into the system.

[0317] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This data is sent to the emotion engine.

[0318] 3. Feedback measures: Providing feedback on the system's response.

[0319] Example: Emotion recognition AI system for medical institutions

[0320] As a specific example, an embodiment of an emotion-recognition conversational artificial intelligence system for medical institutions will be described below:

[0321] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[0322] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[0323] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[0324] Users (doctors and nurses) use conversational AI in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergies," the device sends this input data and the patient's emotional data to the server.

[0325] The server generates an appropriate and emotionally sensitive response based on the user input and emotion data and sends it to the terminal.

[0326] The terminal presents the generated response to the user.

[0327] The server collects and analyzes user-AI dialogue logs and emotional data, and uses the results of this analysis to retrain the model and develop a new dialogue package specialized for medical work.

[0328] In this way, by combining a conversational AI system optimized for a specific industry with an emotion engine, the present invention can provide more precise and attentive responses to users, significantly improving business efficiency for companies.

[0329] The processing flow will be explained below.

[0330] ---

[0331] Step 1:

[0332] The server collects data related to a particular industry from external databases and specialized books.

[0333] Step 2:

[0334] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[0335] Step 3:

[0336] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[0337] Step 4:

[0338] The server incorporates an emotion engine that recognizes the user's emotions, allowing it to analyze the emotion data and reflect it in responses.

[0339] Step 5:

[0340] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[0341] Step 6:

[0342] Users input their company's specific business processes and terminology through the interface.

[0343] Step 7:

[0344] The server collects input data from users and reflects it in the AI ​​model.

[0345] Step 8:

[0346] Users use conversational AI in their daily work, utilizing cameras and microphones to capture emotional data, for example by typing, "What's on the agenda for today's meeting?"

[0347] Step 9:

[0348] The terminal transmits the user's input data and emotion data to the server in real time.

[0349] Step 10:

[0350] The server generates a response based on the user's input and emotion data and sends the response to the terminal.

[0351] Step 11:

[0352] The terminal presents the generated response to the user.

[0353] Step 12:

[0354] The server securely stores the dialogue logs and emotional data between the user and the AI.

[0355] Step 13:

[0356] The server periodically retrieves the accumulated dialogue logs and emotion data and performs data preprocessing.

[0357] Step 14:

[0358] The server analyzes the preprocessed data and sentiment data to identify areas for improvement in the model.

[0359] Step 15:

[0360] The server retrains the conversational AI model based on the analysis results.

[0361] Step 16:

[0362] The server develops new industry-specific dialogue packages.

[0363] Step 17:

[0364] The terminal receives new interaction packages from the server and updates the system.

[0365] Step 18:

[0366] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[0367] Step 19:

[0368] The device sends the user's input data and emotional data to the server, and the conversational AI generates an appropriate response and returns it to the device.

[0369] Step 20:

[0370] The terminal presents the new response to the user.

[0371] Step 21:

[0372] The server continues to store updated dialogue logs and emotional data to help improve the accuracy of the system.

[0373] ---

[0374] These are the specific steps in programming a conversational AI system that combines an emotion engine and is optimized for a specific industry.

[0375] Example 2

[0376] 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."

[0377] Conventional conversational AI systems generate responses only in response to user input and do not take into account the user's emotional state, which can result in a lack of appropriate responses and consideration. Furthermore, generating responses based on the user's emotional state is difficult, which reduces the accuracy and reliability of systems specialized for specific industries.

[0378] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for recognizing and acquiring user emotion data, means for receiving input data and emotion data from the user and analyzing them in real time, means for generating an appropriate response based on the analyzed data, means for presenting the generated response to the user, means for analyzing accumulated dialogue logs and emotion data and retraining the model, means for developing and updating a new industry-specific dialogue package, and means for providing the updated dialogue package to the user. This makes it possible to generate responses according to the user's emotional state, thereby significantly improving business efficiency and reliability in a specific industry.

[0379] "Means for collecting data related to a specific industry" refers to systems or software for automatically collecting information and data related to a specific industry from external databases or specialized books.

[0380] A "means for training an interactive artificial intelligence model" is a method for training an interactive artificial intelligence model using a machine learning algorithm based on collected data.

[0381] "Means for recognizing and acquiring user emotional data" refers to technology that recognizes a user's facial expressions and tone of voice through devices such as a camera or microphone, and acquires their emotional state as data.

[0382] "Means for receiving input data and emotional data from a user and analyzing it in real time" refers to a system in which text, voice data, and emotional data sent by a user are instantly received by a server and analyzed.

[0383] The "means for generating an appropriate response based on the analyzed data" is a method for generating an appropriate response based on user input and emotion data analyzed in real time.

[0384] The "means for presenting the generated response to the user" is a system that displays or audibly conveys the response generated by the server to the user via the terminal.

[0385] "Means for analyzing accumulated dialogue logs and emotional data and retraining a model" refers to a method for analyzing dialogue history and emotional data, and retraining an interactive artificial intelligence model based on the analysis to improve its performance.

[0386] "Means for developing and updating new industry-specific dialogue packages" refers to technology that creates dialogue scenarios and response models optimized for specific industries based on analysis results and new data, and updates the system.

[0387] The "means for providing an updated dialogue package to a user" refers to a method for delivering a newly developed or updated dialogue artificial intelligence package to a user's terminal and making it available for use.

[0388] This invention is a system that provides more accurate responses by combining an emotional engine with a conversational AI system optimized for a specific industry. This system works in conjunction with the server, terminal, and user elements to realize responses based on the user's emotional state.

[0389] Server Features

[0390] The server provides the following main functions:

[0391] 1. Data collection method: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses the PubMed API to obtain medical-related data.

[0392] 2. Model training method: Train an interactive AI model based on the collected data. This method uses machine learning frameworks such as TensorFlow and PyTorch. For example, build an interactive model of the medical treatment process based on medical data.

[0393] 3. Emotion Engine: The server uses emotion recognition software such as Emotion API to analyze the user's facial expressions and tone of voice and obtain emotional data, which is then reflected in the generation of a response.

[0394] 4. Data analysis: Analyzes the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights, thereby improving the efficiency and accuracy of the entire system.

[0395] 5. Package development: Based on the analysis results, new industry-specific dialogue packages are developed and the system is updated. This method is important for generating new dialogue scenarios and response models.

[0396] Device Features

[0397] The terminal provides the following main features:

[0398] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system. For example, in a hospital system, it sets the terminology commonly used by doctors and nurses.

[0399] 2. Data reception and transmission means: Sends input data and emotion data from the user to the server in real time, and presents the response from the server to the user. For this function, for example, the Zoom SDK can be used.

[0400] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system, ensuring that the latest data and response models are always available.

[0401] User operations

[0402] The user does the following:

[0403] 1. Input Method: Conversational AI is used in daily operations to input questions or commands into the system. For example, a doctor might type, "Please tell me Patient X's allergy information." This input can be done via keyboard, voice, or touch panel.

[0404] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This acquired emotion data is sent to the emotion engine on the server.

[0405] 3. Feedback measures: Providing feedback on the system's response. This feedback helps refine and improve the system.

[0406] Specific examples

[0407] For example, an emotion-aware conversational AI system for a medical institution:

[0408] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[0409] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[0410] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[0411] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice.

[0412] Prompt Sentence Examples

[0413] "Please tell me Patient X's allergy information and current emotional state."

[0414] "Please advise on a treatment process that takes into consideration the patient's feelings."

[0415] "Please provide examples of responses for specific medical conditions."

[0416] This invention combines an interactive artificial intelligence system optimized for a specific industry with an emotion engine, which is expected to realize highly accurate responses that are attuned to the user, significantly improving the efficiency and reliability of business operations.

[0417] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0418] Step 1: Data collection

[0419] The server collects data related to a specific industry from external databases and specialized books. The input for this process is a specific keyword or query, and the output is related data. For example, in a system for medical institutions, the server uses the PubMed API to automatically collect literature data based on keywords such as "allergy" and "medical treatment process." The collected data is then used in subsequent model training.

[0420] Step 2: Model training

[0421] The server trains an interactive AI model based on the collected data. The input for this process is the collected data, and the output is a trained interactive AI model. Specifically, it uses machine learning frameworks such as TensorFlow and PyTorch to build and train dialogue scenarios based on medical treatment processes using medical data. The trained model is then used to generate responses to questions from users.

[0422] Step 3: User Input

[0423] The user inputs a question or command into the system. The input for this process is text or voice input from the user, and the output is the input data. For example, a doctor might input "Please tell me about patient X's allergy information" into a terminal. This input data is used in the subsequent emotion recognition and response generation steps.

[0424] Step 4: Emotion Recognition

[0425] The device acquires emotion data from the user's facial expressions and tone of voice via a camera and microphone. The input for this process is the user's visual and audio characteristics, and the output is emotion data. Specifically, OpenCV is used to analyze the doctor's facial expressions captured by the camera to determine whether the user is feeling stressed. The emotion data is sent to the server and used in the subsequent response generation step.

[0426] Step 5: Response Generation

[0427] The server generates an appropriate response based on the user's input data and emotional data. The input for this process is the user's question or command, as well as emotional data, and the output is a response based on that. Specifically, using a generative AI model (e.g., GPT-4 (registered trademark)), if a user inputs "Please tell me Patient X's allergy information," the server generates a response based on the emotional data, such as "Patient X's allergy information is ____. Also, a high stress level has been detected, so special caution is required."

[0428] Step 6: Present the response

[0429] The terminal presents the generated response to the user. The input for this process is the response data received from the server, and the output is the content presented to the user. Specifically, the terminal's display and voice output function are used to notify the doctor, "Patient X's allergy information is ____. Also, it appears that his current stress level is high, so he requires special attention."

[0430] Step 7: Data analysis and model updating

[0431] The server collects and analyzes user-AI dialogue logs and emotional data. The inputs for this process are dialogue logs and emotional data, and the output is the analysis results and an improved dialogue model. Specifically, the accuracy of the model is improved and new insights are extracted based on the dialogue logs and emotional data, and the model is retrained using TensorFlow. The analysis results are used to develop new industry-specific dialogue packages and update the system.

[0432] (Application example 2)

[0433] 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."

[0434] Modern brick-and-mortar stores require effective methods to improve customer shopping experiences. However, conventional methods struggle to grasp customers' emotional states in real time and provide appropriate product recommendations. Furthermore, conversational AI systems generate responses without considering customers' emotions, resulting in low customer satisfaction. Therefore, it is important to provide personalized product recommendations based on customers' emotional states.

[0435] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for customizing the company's business processes and terminology, emotion engine means for recognizing the user's emotional state, analyzing the emotion data, and reflecting the emotion data in responses, and data processing means for providing product suggestions based on the customer's emotional state. This makes it possible to provide personalized product suggestions based on the customer's emotional state in real time.

[0436] "Specific industry" refers to a specific commercial or business sector that specializes in handling data and business processes related to that industry.

[0437] "Data collection means" refers to the means of obtaining information and data related to a specific industry from external databases and materials.

[0438] An "interactive artificial intelligence model" refers to an artificial intelligence algorithm or program that enables natural dialogue with the user.

[0439] A "model training means" is a means for training an interactive artificial intelligence model using collected data to improve its performance.

[0440] "Business process" refers to a series of procedures and flows that a company follows when carrying out its business.

[0441] "Jargon" refers to specialized words and phrases used in a particular industry or field.

[0442] "Customization tools" refers to configuration tools and interfaces that allow a company to adapt the system to its specific business processes and terminology.

[0443] "Input data" refers to information and instructions provided by a user to a system.

[0444] "Interaction log" refers to a record of interaction data between a user and a system.

[0445] "Analytical tools" refers to techniques and methods for analyzing accumulated data and using the results to improve the model.

[0446] "Emotion engine means" refers to technology or algorithms for recognizing the user's emotional state, analyzing that data, and reflecting it in dialogue responses.

[0447] "Data processing means" refers to a mechanism for analyzing acquired data and performing specific processing.

[0448] "Camera and microphone means" refers to a photographic and audio recording device for capturing the customer's facial expressions and tone of voice in real time.

[0449] "Real-time" means near-simultaneous data processing and response.

[0450] "Personalized product proposals" refer to proposals for products and services that are individually customized based on the customer's individual attributes and circumstances.

[0451] The present invention is a system that provides personalized product recommendations using a conversational AI system combined with an emotion engine that recognizes user emotions. This system collects data related to a specific industry, trains a conversational AI model, and customizes the company's business processes and terminology to provide optimized services to customers.

[0452] server

[0453] The server provides the following means:

[0454] 1. Data collection methods: Collect data related to a specific industry from external databases and sources. For example, in the distribution industry, collect and use consumer behavior data.

[0455] 2. Model training method: The collected data is used to train a conversational artificial intelligence model. The model used here is a generative AI model that learns from a large amount of conversational data.

[0456] 3. Emotion engine means: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response. This engine uses the DeepFace library to recognize emotions from facial expressions and tone of voice.

[0457] 4. Data processing means: Processes data to provide product suggestions based on the customer's emotional state. This means obtains product recommendations using external APIs.

[0458] 5. Analysis method: Preprocess the accumulated dialogue logs and perform data analysis, which allows us to retrain the model and improve its accuracy.

[0459] Terminal

[0460] The terminal provides the means to:

[0461] 1. Customization: Provides an interface for setting up a company's business processes and terminology. For example, inputting terminology specific to the distribution industry.

[0462] 2. Input data receiving means: Provides a communication means for transmitting input data from the user to the server in real time.

[0463] 3. Emotion recognition means: Equipped with a camera and microphone to capture the customer's facial expressions and tone of voice in real time. For example, the camera and microphone of a smartphone can be used to recognize the customer's emotions.

[0464] 4. Response providing means: Provides the product suggestions sent from the server to the user. For example, displays the acquired product data on the smartphone screen.

[0465] User

[0466] The user provides the means to:

[0467] 1. Input method: Customers input their questions or opinions using conversational AI. For example, they can input "What's your recommendation today?" through a smartphone app.

[0468] 2. Emotion recognition: Emotion data is acquired from the user's facial expressions and tone of voice via a camera and microphone. This data is sent to the server in real time.

[0469] 3. Feedback measures: Providing feedback on the system's response, e.g., rating satisfaction with a suggested product.

[0470] Specific examples

[0471] For example, if the user is feeling stressed, the system will suggest relaxation-related products, and if the user is smiling, entertainment products that will increase their enjoyment.

[0472] Prompt Sentence Examples

[0473] "The emotion the customer is feeling is stress. Make product recommendations based on this emotion."

[0474] In this way, by using an interactive artificial intelligence system combined with an emotion engine, the present invention can provide personalized services based on the customer's emotional state in real time, thereby improving the shopping experience in physical stores.

[0475] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0476] Step 1:

[0477] The device captures the customer's facial expressions and tone of voice using a camera and microphone. The input is the customer's facial expression data and voice data. This data is collected in real time and saved as image and audio files.

[0478] Step 2:

[0479] The device sends the collected facial expression data and voice data to the server. The input is the image and voice files collected in step 1. Uploading this to the server enables emotion analysis.

[0480] Step 3:

[0481] The server analyzes the facial expression data using the DeepFace library to identify the customer's emotional state. The input is the image file sent in step 2. This data is passed to DeepFace for processing and the output is the emotion analysis result.

[0482] Step 4:

[0483] The server generates a prompt to suggest products appropriate for the customer based on their emotional state and sends a request to the external API. The input is the emotion analysis result obtained in step 3. Based on this, the server generates a prompt such as "The customer is feeling ____. Please suggest products based on this emotion," and sends it to the external API.

[0484] Step 5:

[0485] An external API generates product suggestions based on the prompt sent. The input is the prompt sent in step 4. The API parses it and generates a list of relevant products. The output is the data of the recommended products.

[0486] Step 6:

[0487] The server receives the product suggestions obtained from the external API. The input is the product list data generated in step 5. This data is stored in the server and prepared for transmission to the device.

[0488] Step 7:

[0489] The terminal displays the product proposals received from the server to the customer. The input is the product list data received in step 6. This data is displayed on the smartphone screen, providing specific product information to be proposed to the customer.

[0490] Step 8:

[0491] The user provides feedback on the proposed product. The input is the product information displayed in step 7. The user enters their satisfaction level and opinions about the product and sends them to the server via their terminal.

[0492] Step 9:

[0493] The server analyzes the collected feedback data and uses it to retrain the AI ​​model. The input is the feedback data obtained in step 8. The output is a new and improved AI model.

[0494] In this way, a system is completed that provides personalized product suggestions based on the user's emotional state.

[0495] 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.

[0496] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0497] 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.

[0498] [Second embodiment]

[0499] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0500] 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.

[0501] 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).

[0502] 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.

[0503] 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.

[0504] 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).

[0505] 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.

[0506] 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.

[0507] 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.

[0508] 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.

[0509] 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.

[0510] 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."

[0511] ---

[0512] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[0513] server

[0514] The server provides the following means:

[0515] 1. Data collection method: Collect data related to a specific industry from external databases and specialized books, thereby understanding industry-specific terminology and business processes.

[0516] 2. Model training method: Based on the collected data, an interactive artificial intelligence model is trained, which is built to respond appropriately to industry-specific questions and requests.

[0517] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[0518] 4. Package development method: Based on the analysis results, we develop new industry-specific dialogue packages, thereby providing customized solutions for each company.

[0519] Terminal

[0520] The terminal provides the means to:

[0521] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[0522] 2. Data reception and transmission means: Sends input data from the user to the server in real time, and presents response data from the server to the user.

[0523] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[0524] User

[0525] The user provides the means to:

[0526] 1. Input: Using conversational AI in everyday tasks, you input questions and commands into the system, such as specific requests like "What's on the agenda for today's meeting?" or "Tell me about customer A's past purchase history."

[0527] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[0528] Example: AI system for medical institutions

[0529] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[0530] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[0531] The terminal works in conjunction with the hospital's systems and provides an interface for setting commonly used terms and medical procedures for doctors and nurses.

[0532] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, and the server generates an appropriate response and sends it to the terminal.

[0533] The server collects and analyzes the dialogue logs between users and the AI. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[0534] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[0535] In this way, the present invention provides an interactive artificial intelligence system optimized for a specific industry, which can greatly improve the business efficiency of an enterprise.

[0536] The processing flow will be explained below.

[0537] ---

[0538] Step 1:

[0539] The server collects data related to a particular industry from external databases and specialized books.

[0540] Step 2:

[0541] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[0542] Step 3:

[0543] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[0544] Step 4:

[0545] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[0546] Step 5:

[0547] Users input their company's specific business processes and terminology through the interface.

[0548] Step 6:

[0549] The server collects input data from users and reflects it in the AI ​​model.

[0550] Step 7:

[0551] Users use conversational artificial intelligence in their daily work, for example, by typing, "What's on the agenda for today's meetings?"

[0552] Step 8:

[0553] The terminal transmits the user's input to the server in real time.

[0554] Step 9:

[0555] The server generates a response based on the user's input and sends the response to the terminal.

[0556] Step 10:

[0557] The terminal presents the generated response to the user.

[0558] Step 11:

[0559] The server securely stores the dialogue logs between the user and the AI.

[0560] Step 12:

[0561] The server periodically retrieves the accumulated dialogue logs and performs data preprocessing.

[0562] Step 13:

[0563] The server analyzes the pre-processed data and identifies areas for improvement in the model.

[0564] Step 14:

[0565] The server retrains the conversational AI model based on the analysis results.

[0566] Step 15:

[0567] The server develops new industry-specific dialogue packages.

[0568] Step 16:

[0569] The terminal receives new interaction packages from the server and updates the system.

[0570] Step 17:

[0571] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[0572] Step 18:

[0573] The device sends the user's input to the server, and the conversational AI generates a response and returns it to the device.

[0574] Step 19:

[0575] The terminal presents the new response to the user.

[0576] Step 20:

[0577] The server continues to store updated interaction logs to help improve the accuracy of the system.

[0578] ---

[0579] These are the specific steps in programming an interactive AI system optimized for a specific industry.

[0580] Example 1

[0581] 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."

[0582] Conventional conversational AI systems have struggled to consistently provide everything from data collection to model training and customization, dialogue log accumulation and analysis, and automatic system updates to meet the specialized requirements of specific industries. As a result, they have been unable to provide highly accurate responses tailored to each company's needs. While this is expected to improve business efficiency, it actually places a heavy burden on users and lacks convenience. In particular, there is a need for appropriate and rapid response in the process of generating real-time responses, refining models, and updating dialogue packages based on feedback.

[0583] 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.

[0584] In this invention, the server includes a means for collecting data related to a specific industry, a means for training an interactive AI model using the collected data, and a means for generating responses to users. This makes it possible to collect data suited to the specialized needs of each company and generate highly accurate responses in real time. Furthermore, by supporting customization, analysis of dialogue logs, model retraining based on feedback, and automatic updates of dialogue packages, it is possible to improve business efficiency and reduce the burden on users.

[0585] "Means for collecting data related to a specific industry" refers to methods or devices for obtaining information related to a specific industry from external databases or specialized books.

[0586] "Means for training an interactive AI model using collected data" refers to a method or device for building an interactive AI model using a machine learning algorithm based on acquired data and optimizing the model.

[0587] "Means for customizing a company's business processes and terminology" refers to methods and devices that provide interfaces and configuration functions for reflecting the unique business procedures and terminology of each company within the system.

[0588] "Means for receiving input data from a user and transmitting it to a server in real time" refers to a communication protocol or device for instantly receiving data input by a user into the system and transmitting it to a server.

[0589] A "means for generating a response to a user" is a method or device that utilizes an interactive artificial intelligence model to generate an appropriate reply based on input data from a user.

[0590] "Means for analyzing dialogue logs and retraining models" refers to a method or device that analyzes dialogue history with a user and retrains the model based on this to improve the accuracy of the dialogue-based artificial intelligence model.

[0591] "Means for developing and updating new industry-specific dialogue packages" refers to methods and devices for creating new dialogue packages tailored to specific industries based on analysis results and feedback, and for keeping the system up to date.

[0592] "Means for delivering updated dialogue packages to terminals and automatically updating the system" refers to a method or device for applying the latest dialogue packages delivered from a server to terminals and automatically updating the system.

[0593] MODE FOR CARRYING OUT THE INVENTION

[0594] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[0595] server

[0596] The server provides the following means:

[0597] 1. A means of collecting data related to a specific industry: The server obtains information related to a specific industry from external databases or specialized books. For example, in the medical industry, data is collected from PubMed and other medical databases. During this process, the data is acquired and analyzed using Python's requests library, BeautifulSoup, and Scrapy.

[0598] 2. Means for training a conversational AI model using the collected data: The server uses the collected data to train a conversational AI model using TensorFlow or PyTorch. Specifically, it uses natural language processing (NLP) techniques to build a model that can appropriately respond to questions and requests specific to the industry.

[0599] 3. Means for generating a response to the user: The server uses an interactive artificial intelligence model to generate an appropriate response based on the input data from the user. For example, if a doctor inputs "Please tell me about patient X's allergies," the server will query the electronic medical record database and generate an appropriate response from the results.

[0600] 4. Analyzing the dialogue log and retraining the model: The server analyzes the dialogue history with the user and retrains the dialogue AI model to improve its accuracy. It uses Python's pandas and NumPy to analyze the log data and retrains the model based on the analysis results.

[0601] 5. Means for developing and updating new industry-specific dialogue packages: The server creates new dialogue packages tailored to specific industries based on analysis results and feedback, and updates the system to keep it up to date.

[0602] Terminal

[0603] The terminal provides the means to:

[0604] 1. A means to customize a company's business processes and terminology: The terminal provides an interface that reflects a company's specific business procedures and terminology within the system. For example, a customized interface can be built using a front-end framework such as React or Vue.js.

[0605] 2. Means for receiving input data from the user and transmitting it to the server in real time: The terminal uses a communication protocol (e.g., REST API or WebSocket) to transmit data input by the user to the server in real time.

[0606] 3. Means for delivering updated dialogue packages to terminals and automatically updating the system: The terminals receive the latest dialogue packages delivered from the server and automatically update the system, allowing users to always interact with the latest information.

[0607] User

[0608] The user provides the means to:

[0609] 1. Input means: A user uses an interactive AI system to input specific questions or commands into the system. For example, a specific request such as "Please tell me the allergy information of patient X" is input into a terminal.

[0610] 2. Feedback: The user provides feedback on the system's response. They evaluate whether the response is accurate or there is room for improvement, and send the feedback to the server via their device. This feedback is used to retrain the model and improve the system.

[0611] Example: AI system for medical institutions

[0612] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[0613] The server collects data on medical terminology and medical procedures from medical databases and uses TensorFlow to train an interactive AI model. For example, it can collect research paper data from PubMed via an API and use it to build a neural network model.

[0614] The device connects to the hospital's electronic medical record system and provides an interface for doctors and nurses to set commonly used terms and medical procedures. This is achieved through a web application written in React.

[0615] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, which retrieves the data from the electronic medical record, generates an appropriate response, and returns it to the terminal.

[0616] The server collects dialogue logs between users and the AI ​​and analyzes them using Python's pandas. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[0617] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[0618] Examples of prompt statements

[0619] Below are some example prompts from an AI model for healthcare:

[0620] "I'd like to see Patient X's most recent medical records. What procedures were performed?"

[0621] By inputting this prompt into a generative AI model, it is possible to extract appropriate medical information. Furthermore, it can also suggest improvements to the prompt based on the results of analyzing actual dialogue logs.

[0622] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0623] Step 1: Data Collection (Server)

[0624] Specific operation: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses Python's requests library to send an HTTP request to the PubMed API and parses the returned JSON response.

[0625] Input: Data related to a specific industry (e.g., PubMed API)

[0626] Data processing: HTTP request sending, response analysis

[0627] Output: Retrieved data (JSON format)

[0628] Step 2: Model training (server)

[0629] Specific operation: The server uses the collected data to train an interactive artificial intelligence model. For example, it uses the TensorFlow library to build a neural network model and supplies training data to train the model.

[0630] Input: Collected data (JSON format)

[0631] Data processing: Data preprocessing (cleaning, tokenization), model building, training

[0632] Output: A trained interactive AI model

[0633] Step 3: Interface Settings (Device)

[0634] Specific operation: The terminal provides an interface for inputting a company's business processes and terminology into the system. The terminal uses React to build a customized interface, allowing users to input company-specific information.

[0635] Input: Company business processes and terminology (user input)

[0636] Data processing: Interface construction

[0637] Output: Customized configuration information

[0638] Step 4: User Input (User)

[0639] Specific Actions: A user (e.g., a doctor or nurse) uses a conversational AI system to input specific questions or commands into the system, and transmits data to the system using text input or voice recognition technology (e.g., Google Speech-to-Text API).

[0640] Input: User questions or commands (text or voice)

[0641] Data processing: Voice recognition (when inputting voice)

[0642] Output: User input data (text format)

[0643] Step 5: Send data (terminal)

[0644] Specific operation: The terminal sends user input data to the server in real time. To do this, the terminal uses REST API or WebSocket to send data with low latency.

[0645] Input: User input data (text format)

[0646] Data processing: Data encoding, HTTP POST request or WebSocket transmission

[0647] Output: Data sent to the server

[0648] Step 6: Response Generation (Server)

[0649] Specific operation: The server generates a response using an interactive artificial intelligence model based on the received user input data. The server sends a query to the electronic medical record database and generates an appropriate response from the results.

[0650] Input: User input data (text format)

[0651] Data processing: query generation, database interrogation, response generation

[0652] Output: Generated response data (text format)

[0653] Step 7: Response presentation (terminal)

[0654] Specific operation: The terminal presents the response data returned from the server to the user. The terminal dynamically updates the web page and displays the response in a user-friendly format.

[0655] Input: Response data from the server (text format)

[0656] Data processing: Data decoding, display format conversion

[0657] Output: The response presented to the user

[0658] Step 8: Feedback (User)

[0659] Specific operation: The user provides feedback on the system's response by entering an evaluation and suggestions for improvement through the terminal and sending the feedback to the server.

[0660] Input: User feedback (text format)

[0661] Data processing: Evaluation data input

[0662] Output: Feedback sent to the server

[0663] Step 9: Log analysis and model update (server)

[0664] How it works: The server analyzes the collected interaction logs and feedback. It uses Python's pandas and NumPy to extract insights from the log data and retrain the model.

[0665] Input: Dialogue logs, feedback data

[0666] Data processing: log analysis, data preprocessing, model retraining

[0667] Output: Updated interactive artificial intelligence model

[0668] Step 10: Update packages (device)

[0669] Specific operation: The terminal receives new industry-specific dialogue packages distributed by the server and automatically updates the system, so that the latest dialogue models and functions are applied to the terminal.

[0670] Input: New dialogue package

[0671] Data processing: Package download, update application

[0672] Output: Updated system

[0673] (Application example 1)

[0674] 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."

[0675] In modern business, there is a growing demand for conversational AI systems to efficiently handle specific industries. In particular, the food delivery industry needs ways to streamline order taking, delivery tracking, and customer support. However, performing these tasks manually is time-consuming, labor-intensive, and prone to errors. Therefore, the present invention aims to provide an industry-specific conversational AI system to solve these problems.

[0676] 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.

[0677] In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive AI model using the collected data, means for customizing a company's business processes and terminology, means for receiving input data from a user and accumulating an interaction log, means for analyzing the accumulated interaction log and retraining the model, means for developing and updating a new industry-specific interaction package, means for providing the updated interaction package to the user, means for converting user voice input into text, means for analyzing user order information using the interactive AI model and providing menu options, and means for tracking delivery status for a specific order and providing the user with an estimated delivery time. This makes it possible to streamline a series of operations, from order taking to delivery tracking and customer support, even in the food delivery industry.

[0678] "Specific industry" refers to all businesses related to a specific business area or field.

[0679] "Data collection" refers to the process of obtaining information relevant to a particular industry.

[0680] "Conversational artificial intelligence model" refers to a machine learning model that is trained to respond appropriately to user input in natural language.

[0681] "A company's business processes" refers to the set of procedures and methods by which a company carries out its daily business operations.

[0682] "Jargon" refers to words and phrases specific to a particular industry or field.

[0683] "Customization" refers to the process of tailoring a system or service to meet specific needs and requirements.

[0684] "Input Data" refers to information or instructions provided by a user.

[0685] "Dialogue log" refers to a record of all conversations exchanged between a user and an interactive AI system.

[0686] "Data analysis" refers to the process of analyzing accumulated data and extracting meaningful information and patterns.

[0687] "Model retraining" refers to the process of retraining an existing machine learning model based on new data or analytical results.

[0688] "Dialogue package" refers to a software module that forms part of a dialogue-based artificial intelligence system optimized for a specific industry.

[0689] "Voice input" refers to the process of capturing user spoken words into the system.

[0690] "Menu options" refer to selectable products and services offered to a user.

[0691] "Delivery status" refers to the progress of an ordered item until it is delivered.

[0692] "Expected Delivery Time" means the estimated time it will take for an ordered item to be delivered.

[0693] The present invention provides an interactive artificial intelligence system optimized for a specific industry, streamlining order acceptance, delivery tracking, and customer support in the food delivery industry. This system is configured to operate in cooperation with each element: server, terminal, and user.

[0694] server

[0695] The server provides the following means:

[0696] 1. Data collection method: Collect information related to a specific industry from external databases to understand industry-specific terminology and business processes.

[0697] 2. Model training method: Based on the collected data, a conversational artificial intelligence model (generative AI model) is trained, which builds a model that can respond appropriately to industry-specific questions and requests.

[0698] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[0699] 4. Package development tools: Based on the analysis results, we will develop new industry-specific interactive packages to optimize order taking, delivery tracking, and customer support in the food delivery industry.

[0700] Terminal

[0701] The terminal provides the means to:

[0702] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[0703] 2. Data reception and transmission means: Data input from the user (e.g., order details or questions) is transmitted to the server in real time, and response data from the server is presented to the user.

[0704] 3. Voice input means: Provides a voice recognition function to convert voice input into text.

[0705] 4. Package update means: Receive new industry-specific dialogue packages (generative AI models) distributed from the server and update the system.

[0706] User

[0707] The user provides the means to:

[0708] 1. Input method: Food delivery orders and inquiries are input into the conversational AI system. For example, you can say, "I'd like to order one Margherita pizza and one Coke Zero" by voice.

[0709] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[0710] Hardware and software used

[0711] Hardware: Cloud servers (AWS, Google Cloud, etc.), smartphones (iOS, Android, etc.)

[0712] Software: Generative AI model (facebook / blenderbot-400M-distill), framework (Flask), speech recognition API (Google Cloud Speech-to-Text), database (MongoDB)

[0713] Data processing and calculation

[0714] Speech-to-text conversion: Uses the Google Cloud Speech-to-Text API to convert speech to text, allowing you to capture user speech as text.

[0715] Dialogue model response generation: A generative AI model trained on collected data is used to generate appropriate responses to user input. For example, if a user inputs, "I'd like to order one Margherita pizza and one Coke Zero," the conversational AI system will provide corresponding menu options.

[0716] Delivery status tracking and provision: Real-time delivery status information for specific orders is tracked and provided to users. For example, if a user requests "What is the delivery status of order number 12345?", the current delivery status and estimated delivery time will be provided.

[0717] Specific examples

[0718] Order Taker: "I'd like to order one Margherita pizza and one Coke Zero."

[0719] Delivery tracking: "What's the delivery status of order number 12345?"

[0720] Customer Support: "Why can't I use my coupon code?"

[0721] Prompt Sentence Examples

[0722] Enter your order: "I'd like to order one Margherita pizza and one Coke Zero."

[0723] Delivery tracking: "What's the delivery status of order number 12345?"

[0724] Support Enquiry: "Why can't I use my coupon code?"

[0725] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0726] Step 1:

[0727] Users voice-order food delivery through a smartphone application, where the voice input is captured.

[0728] Input: User speaks "I would like to order one Margherita pizza and one Coke Zero."

[0729] Output: Audio input data

[0730] Step 2:

[0731] The device converts voice input into text using the Google Cloud Speech-to-Text API.

[0732] Input: Voice input data

[0733] Output: Text data "I would like to order one Margherita pizza and one Coke Zero."

[0734] Specific operation: Call the voice recognition API and obtain the voice data as text data.

[0735] Step 3:

[0736] The device sends text data to the server, which uses a generative AI model (facebook / blenderbot-400M-distill) to generate an appropriate response to the user's order.

[0737] Input: Text data "I would like to order one Margherita pizza and one Coke Zero."

[0738] Output: Response data containing menu options

[0739] Specific operation: Text data is input into an interactive artificial intelligence model and response text is generated.

[0740] Step 4:

[0741] The server transmits the generated response data to the terminal, which then presents it to the user.

[0742] Input: Response data "Would you like one Margherita pizza and one Coke Zero?"

[0743] Output: The response text to display to the user

[0744] Specific behavior: Display the response text on the device's user interface.

[0745] Step 5:

[0746] The user provides confirmation, for example, a voice or text reply of "Yes, please."

[0747] Input: User response "Yes, please"

[0748] Output: Input data for verification

[0749] Step 6:

[0750] The terminal performs voice recognition again and sends the result as text data to the server. The server receives this confirmation data and processes it to confirm the order.

[0751] Input: Input data for confirmation "Yes, please"

[0752] Output: Order confirmation data

[0753] Specific operation: Converts speech into text and sends the data to the server to confirm the order. The server stores the order information in a database.

[0754] Step 7:

[0755] After the order is confirmed, the server generates a tracking ID for tracking the delivery status and sends it to the terminal. The terminal displays the tracking ID to the user.

[0756] Input: Order confirmation data

[0757] Output: Tracking ID

[0758] Specific operation: A tracking ID is generated based on the order information and sent to the terminal, which then displays the tracking ID on the user interface.

[0759] Step 8:

[0760] A user queries the application to check the delivery status of an order, entering something like "What is the delivery status of order number 12345?"

[0761] Input: User inquiry: "What is the delivery status of order number 12345?"

[0762] Output: Delivery tracking request data

[0763] Step 9:

[0764] The server tracks the delivery status, obtains the status in real time, and transmits the data to the terminal, which then presents the status data to the user.

[0765] Input: Delivery tracking request data

[0766] Output: Delivery status data "Current delivery status is on its way and will arrive in 15 minutes."

[0767] Specific operation: Obtain the current status from the delivery tracking system and send the information to the terminal.

[0768] Step 10:

[0769] The order is completed when the user receives it. The user provides feedback to the system, for example, by typing "The item arrived safely."

[0770] Input: User feedback "The item arrived safely."

[0771] Output: Feedback data

[0772] Specific operation: Feedback data is sent to the server and used to improve the system.

[0773] 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.

[0774] ---

[0775] This invention is a system that provides more sophisticated responses by combining a conversational AI system optimized for a specific industry with an emotion engine that recognizes the user's emotions. This system is configured so that the server, terminal, and user elements work together and the emotion engine is used to realize responses based on the user's emotional state.

[0776] server

[0777] The server provides the following means:

[0778] 1. Data collection methods: Data related to a specific industry are collected from external databases and specialized books.

[0779] 2. Model training method: Train an interactive artificial intelligence model based on the collected data.

[0780] 3. Emotion engine: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response.

[0781] 4. Data analysis: Analyze the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights.

[0782] 5. Package development method: Based on the analysis results, new industry-specific dialogue packages will be developed.

[0783] Terminal

[0784] The terminal provides the means to:

[0785] 1. Customization methods: Provide an interface for inputting a company's business processes and terminology into the system.

[0786] 2. Data reception and transmission means: Input data and emotion data from the user are transmitted to the server in real time, and response data from the server is presented to the user.

[0787] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[0788] User

[0789] The user provides the means to:

[0790] 1. Input means: Use conversational artificial intelligence in your daily work by inputting questions and commands into the system.

[0791] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This data is sent to the emotion engine.

[0792] 3. Feedback measures: Providing feedback on the system's response.

[0793] Example: Emotion recognition AI system for medical institutions

[0794] As a specific example, an embodiment of an emotion-recognition conversational artificial intelligence system for medical institutions will be described below:

[0795] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[0796] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[0797] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[0798] Users (doctors and nurses) use conversational AI in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergies," the device sends this input data and the patient's emotional data to the server.

[0799] The server generates an appropriate and emotionally sensitive response based on the user input and emotion data and sends it to the terminal.

[0800] The terminal presents the generated response to the user.

[0801] The server collects and analyzes user-AI dialogue logs and emotional data, and uses the results of this analysis to retrain the model and develop a new dialogue package specialized for medical work.

[0802] In this way, by combining a conversational AI system optimized for a specific industry with an emotion engine, the present invention can provide more precise and attentive responses to users, significantly improving business efficiency for companies.

[0803] The processing flow will be explained below.

[0804] ---

[0805] Step 1:

[0806] The server collects data related to a particular industry from external databases and specialized books.

[0807] Step 2:

[0808] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[0809] Step 3:

[0810] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[0811] Step 4:

[0812] The server incorporates an emotion engine that recognizes the user's emotions, allowing it to analyze the emotion data and reflect it in responses.

[0813] Step 5:

[0814] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[0815] Step 6:

[0816] Users input their company's specific business processes and terminology through the interface.

[0817] Step 7:

[0818] The server collects input data from users and reflects it in the AI ​​model.

[0819] Step 8:

[0820] Users use conversational AI in their daily work, utilizing cameras and microphones to capture emotional data, for example by typing, "What's on the agenda for today's meeting?"

[0821] Step 9:

[0822] The terminal transmits the user's input data and emotion data to the server in real time.

[0823] Step 10:

[0824] The server generates a response based on the user's input and emotion data and sends the response to the terminal.

[0825] Step 11:

[0826] The terminal presents the generated response to the user.

[0827] Step 12:

[0828] The server securely stores the dialogue logs and emotional data between the user and the AI.

[0829] Step 13:

[0830] The server periodically retrieves the accumulated dialogue logs and emotion data and performs data preprocessing.

[0831] Step 14:

[0832] The server analyzes the preprocessed data and sentiment data to identify areas for improvement in the model.

[0833] Step 15:

[0834] The server retrains the conversational AI model based on the analysis results.

[0835] Step 16:

[0836] The server develops new industry-specific dialogue packages.

[0837] Step 17:

[0838] The terminal receives new interaction packages from the server and updates the system.

[0839] Step 18:

[0840] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[0841] Step 19:

[0842] The device sends the user's input data and emotional data to the server, and the conversational AI generates an appropriate response and returns it to the device.

[0843] Step 20:

[0844] The terminal presents the new response to the user.

[0845] Step 21:

[0846] The server continues to store updated dialogue logs and emotional data to help improve the accuracy of the system.

[0847] ---

[0848] These are the specific steps in programming a conversational AI system that combines an emotion engine and is optimized for a specific industry.

[0849] Example 2

[0850] 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."

[0851] Conventional conversational AI systems generate responses only in response to user input and do not take into account the user's emotional state, which can result in a lack of appropriate responses and consideration. Furthermore, generating responses based on the user's emotional state is difficult, which reduces the accuracy and reliability of systems specialized for specific industries.

[0852] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for recognizing and acquiring user emotion data, means for receiving input data and emotion data from the user and analyzing them in real time, means for generating an appropriate response based on the analyzed data, means for presenting the generated response to the user, means for analyzing accumulated dialogue logs and emotion data and retraining the model, means for developing and updating a new industry-specific dialogue package, and means for providing the updated dialogue package to the user. This makes it possible to generate responses according to the user's emotional state, thereby significantly improving business efficiency and reliability in a specific industry.

[0853] "Means for collecting data related to a specific industry" refers to systems or software for automatically collecting information and data related to a specific industry from external databases or specialized books.

[0854] A "means for training an interactive artificial intelligence model" is a method for training an interactive artificial intelligence model using a machine learning algorithm based on collected data.

[0855] "Means for recognizing and acquiring user emotional data" refers to technology that recognizes a user's facial expressions and tone of voice through devices such as a camera or microphone, and acquires their emotional state as data.

[0856] "Means for receiving input data and emotional data from a user and analyzing it in real time" refers to a system in which text, voice data, and emotional data sent by a user are instantly received by a server and analyzed.

[0857] The "means for generating an appropriate response based on the analyzed data" is a method for generating an appropriate response based on user input and emotion data analyzed in real time.

[0858] The "means for presenting the generated response to the user" is a system that displays or audibly conveys the response generated by the server to the user via the terminal.

[0859] "Means for analyzing accumulated dialogue logs and emotional data and retraining a model" refers to a method for analyzing dialogue history and emotional data, and retraining an interactive artificial intelligence model based on the analysis to improve its performance.

[0860] "Means for developing and updating new industry-specific dialogue packages" refers to technology that creates dialogue scenarios and response models optimized for specific industries based on analysis results and new data, and updates the system.

[0861] The "means for providing an updated dialogue package to a user" refers to a method for delivering a newly developed or updated dialogue artificial intelligence package to a user's terminal and making it available for use.

[0862] This invention is a system that provides more accurate responses by combining an emotional engine with a conversational AI system optimized for a specific industry. This system works in conjunction with the server, terminal, and user elements to realize responses based on the user's emotional state.

[0863] Server Features

[0864] The server provides the following main functions:

[0865] 1. Data collection method: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses the PubMed API to obtain medical-related data.

[0866] 2. Model training method: Train an interactive AI model based on the collected data. This method uses machine learning frameworks such as TensorFlow and PyTorch. For example, build an interactive model of the medical treatment process based on medical data.

[0867] 3. Emotion Engine: The server uses emotion recognition software such as Emotion API to analyze the user's facial expressions and tone of voice and obtain emotional data, which is then reflected in the generation of a response.

[0868] 4. Data analysis: Analyzes the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights, thereby improving the efficiency and accuracy of the entire system.

[0869] 5. Package development: Based on the analysis results, new industry-specific dialogue packages are developed and the system is updated. This method is important for generating new dialogue scenarios and response models.

[0870] Device Features

[0871] The terminal provides the following main features:

[0872] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system. For example, in a hospital system, it sets the terminology commonly used by doctors and nurses.

[0873] 2. Data reception and transmission means: Sends input data and emotion data from the user to the server in real time, and presents the response from the server to the user. For this function, for example, the Zoom SDK can be used.

[0874] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system, ensuring that the latest data and response models are always available.

[0875] User operations

[0876] The user does the following:

[0877] 1. Input Method: Conversational AI is used in daily operations to input questions or commands into the system. For example, a doctor might type, "Please tell me Patient X's allergy information." This input can be done via keyboard, voice, or touch panel.

[0878] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This acquired emotion data is sent to the emotion engine on the server.

[0879] 3. Feedback measures: Providing feedback on the system's response. This feedback helps refine and improve the system.

[0880] Specific examples

[0881] For example, an emotion-aware conversational AI system for a medical institution:

[0882] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[0883] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[0884] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[0885] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice.

[0886] Prompt Sentence Examples

[0887] "Please tell me Patient X's allergy information and current emotional state."

[0888] "Please advise on a treatment process that takes into consideration the patient's feelings."

[0889] "Please provide examples of responses for specific medical conditions."

[0890] This invention combines an interactive artificial intelligence system optimized for a specific industry with an emotion engine, which is expected to realize highly accurate responses that are attuned to the user, significantly improving the efficiency and reliability of business operations.

[0891] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0892] Step 1: Data collection

[0893] The server collects data related to a specific industry from external databases and specialized books. The input for this process is a specific keyword or query, and the output is related data. For example, in a system for medical institutions, the server uses the PubMed API to automatically collect literature data based on keywords such as "allergy" and "medical treatment process." The collected data is then used in subsequent model training.

[0894] Step 2: Model training

[0895] The server trains an interactive AI model based on the collected data. The input for this process is the collected data, and the output is a trained interactive AI model. Specifically, it uses machine learning frameworks such as TensorFlow and PyTorch to build and train dialogue scenarios based on medical treatment processes using medical data. The trained model is then used to generate responses to questions from users.

[0896] Step 3: User Input

[0897] The user inputs a question or command into the system. The input for this process is text or voice input from the user, and the output is the input data. For example, a doctor might input "Please tell me about patient X's allergy information" into a terminal. This input data is used in the subsequent emotion recognition and response generation steps.

[0898] Step 4: Emotion Recognition

[0899] The device acquires emotion data from the user's facial expressions and tone of voice via a camera and microphone. The input for this process is the user's visual and audio characteristics, and the output is emotion data. Specifically, OpenCV is used to analyze the doctor's facial expressions captured by the camera to determine whether the user is feeling stressed. The emotion data is sent to the server and used in the subsequent response generation step.

[0900] Step 5: Response Generation

[0901] The server generates an appropriate response based on the user's input data and emotional data. The input for this process is the user's question or command, as well as emotional data, and the output is a response based on that. Specifically, using a generative AI model (e.g., GPT-4), if a user inputs "Tell me about patient X's allergies," the server generates a response based on the emotional data, such as "Patient X's allergy information is ____. Also, a high stress level has been detected, so special care is required."

[0902] Step 6: Present the response

[0903] The terminal presents the generated response to the user. The input for this process is the response data received from the server, and the output is the content presented to the user. Specifically, the terminal's display and voice output function are used to notify the doctor, "Patient X's allergy information is ____. Also, it appears that his current stress level is high, so he requires special attention."

[0904] Step 7: Data analysis and model updating

[0905] The server collects and analyzes user-AI dialogue logs and emotional data. The inputs for this process are dialogue logs and emotional data, and the output is the analysis results and an improved dialogue model. Specifically, the accuracy of the model is improved and new insights are extracted based on the dialogue logs and emotional data, and the model is retrained using TensorFlow. The analysis results are used to develop new industry-specific dialogue packages and update the system.

[0906] (Application example 2)

[0907] 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."

[0908] Modern brick-and-mortar stores require effective methods to improve customer shopping experiences. However, conventional methods struggle to grasp customers' emotional states in real time and provide appropriate product recommendations. Furthermore, conversational AI systems generate responses without considering customers' emotions, resulting in low customer satisfaction. Therefore, it is important to provide personalized product recommendations based on customers' emotional states.

[0909] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for customizing the company's business processes and terminology, emotion engine means for recognizing the user's emotional state, analyzing the emotion data, and reflecting the emotion data in responses, and data processing means for providing product suggestions based on the customer's emotional state. This makes it possible to provide personalized product suggestions based on the customer's emotional state in real time.

[0910] "Specific industry" refers to a specific commercial or business sector that specializes in handling data and business processes related to that industry.

[0911] "Data collection means" refers to the means of obtaining information and data related to a specific industry from external databases and materials.

[0912] An "interactive artificial intelligence model" refers to an artificial intelligence algorithm or program that enables natural dialogue with the user.

[0913] A "model training means" is a means for training an interactive artificial intelligence model using collected data to improve its performance.

[0914] "Business process" refers to a series of procedures and flows that a company follows when carrying out its business.

[0915] "Jargon" refers to specialized words and phrases used in a particular industry or field.

[0916] "Customization tools" refers to configuration tools and interfaces that allow a company to adapt the system to its specific business processes and terminology.

[0917] "Input data" refers to information and instructions provided by a user to a system.

[0918] "Interaction log" refers to a record of interaction data between a user and a system.

[0919] "Analytical tools" refers to techniques and methods for analyzing accumulated data and using the results to improve the model.

[0920] "Emotion engine means" refers to technology or algorithms for recognizing the user's emotional state, analyzing that data, and reflecting it in dialogue responses.

[0921] "Data processing means" refers to a mechanism for analyzing acquired data and performing specific processing.

[0922] "Camera and microphone means" refers to a photographic and audio recording device for capturing the customer's facial expressions and tone of voice in real time.

[0923] "Real-time" means near-simultaneous data processing and response.

[0924] "Personalized product proposals" refer to proposals for products and services that are individually customized based on the customer's individual attributes and circumstances.

[0925] The present invention is a system that provides personalized product recommendations using a conversational AI system combined with an emotion engine that recognizes user emotions. This system collects data related to a specific industry, trains a conversational AI model, and customizes the company's business processes and terminology to provide optimized services to customers.

[0926] server

[0927] The server provides the following means:

[0928] 1. Data collection methods: Collect data related to a specific industry from external databases and sources. For example, in the distribution industry, collect and use consumer behavior data.

[0929] 2. Model training method: The collected data is used to train a conversational artificial intelligence model. The model used here is a generative AI model that learns from a large amount of conversational data.

[0930] 3. Emotion engine means: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response. This engine uses the DeepFace library to recognize emotions from facial expressions and tone of voice.

[0931] 4. Data processing means: Processes data to provide product suggestions based on the customer's emotional state. This means obtains product recommendations using external APIs.

[0932] 5. Analysis method: Preprocess the accumulated dialogue logs and perform data analysis, which allows us to retrain the model and improve its accuracy.

[0933] Terminal

[0934] The terminal provides the means to:

[0935] 1. Customization: Provides an interface for setting up a company's business processes and terminology. For example, inputting terminology specific to the distribution industry.

[0936] 2. Input data receiving means: Provides a communication means for transmitting input data from the user to the server in real time.

[0937] 3. Emotion recognition means: Equipped with a camera and microphone to capture the customer's facial expressions and tone of voice in real time. For example, the camera and microphone of a smartphone can be used to recognize the customer's emotions.

[0938] 4. Response providing means: Provides the product suggestions sent from the server to the user. For example, displays the acquired product data on the smartphone screen.

[0939] User

[0940] The user provides the means to:

[0941] 1. Input method: Customers input their questions or opinions using conversational AI. For example, they can input "What's your recommendation today?" through a smartphone app.

[0942] 2. Emotion recognition: Emotion data is acquired from the user's facial expressions and tone of voice via a camera and microphone. This data is sent to the server in real time.

[0943] 3. Feedback measures: Providing feedback on the system's response, e.g., rating satisfaction with a suggested product.

[0944] Specific examples

[0945] For example, if the user is feeling stressed, the system will suggest relaxation-related products, and if the user is smiling, entertainment products that will increase their enjoyment.

[0946] Prompt Sentence Examples

[0947] "The emotion the customer is feeling is stress. Make product recommendations based on this emotion."

[0948] In this way, by using an interactive artificial intelligence system combined with an emotion engine, the present invention can provide personalized services based on the customer's emotional state in real time, thereby improving the shopping experience in physical stores.

[0949] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0950] Step 1:

[0951] The device captures the customer's facial expressions and tone of voice using a camera and microphone. The input is the customer's facial expression data and voice data. This data is collected in real time and saved as image and audio files.

[0952] Step 2:

[0953] The device sends the collected facial expression data and voice data to the server. The input is the image and voice files collected in step 1. Uploading this to the server enables emotion analysis.

[0954] Step 3:

[0955] The server analyzes the facial expression data using the DeepFace library to identify the customer's emotional state. The input is the image file sent in step 2. This data is passed to DeepFace for processing and the output is the emotion analysis result.

[0956] Step 4:

[0957] The server generates a prompt to suggest products appropriate for the customer based on their emotional state and sends a request to the external API. The input is the emotion analysis result obtained in step 3. Based on this, the server generates a prompt such as "The customer is feeling ____. Please suggest products based on this emotion," and sends it to the external API.

[0958] Step 5:

[0959] An external API generates product suggestions based on the prompt sent. The input is the prompt sent in step 4. The API parses it and generates a list of relevant products. The output is the data of the recommended products.

[0960] Step 6:

[0961] The server receives the product suggestions obtained from the external API. The input is the product list data generated in step 5. This data is stored in the server and prepared for transmission to the device.

[0962] Step 7:

[0963] The terminal displays the product proposals received from the server to the customer. The input is the product list data received in step 6. This data is displayed on the smartphone screen, providing specific product information to be proposed to the customer.

[0964] Step 8:

[0965] The user provides feedback on the proposed product. The input is the product information displayed in step 7. The user enters their satisfaction level and opinions about the product and sends them to the server via their terminal.

[0966] Step 9:

[0967] The server analyzes the collected feedback data and uses it to retrain the AI ​​model. The input is the feedback data obtained in step 8. The output is a new and improved AI model.

[0968] In this way, a system is completed that provides personalized product suggestions based on the user's emotional state.

[0969] 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.

[0970] 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.

[0971] 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.

[0972] [Third embodiment]

[0973] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0974] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0975] 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).

[0976] 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.

[0977] 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.

[0978] 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).

[0979] 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.

[0980] 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.

[0981] 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.

[0982] 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.

[0983] 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.

[0984] 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."

[0985] ---

[0986] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[0987] server

[0988] The server provides the following means:

[0989] 1. Data collection method: Collect data related to a specific industry from external databases and specialized books, thereby understanding industry-specific terminology and business processes.

[0990] 2. Model training method: Based on the collected data, an interactive artificial intelligence model is trained, which is built to respond appropriately to industry-specific questions and requests.

[0991] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[0992] 4. Package development method: Based on the analysis results, we develop new industry-specific dialogue packages, thereby providing customized solutions for each company.

[0993] Terminal

[0994] The terminal provides the means to:

[0995] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[0996] 2. Data reception and transmission means: Sends input data from the user to the server in real time, and presents response data from the server to the user.

[0997] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[0998] User

[0999] The user provides the means to:

[1000] 1. Input: Using conversational AI in everyday tasks, you input questions and commands into the system, such as specific requests like "What's on the agenda for today's meeting?" or "Tell me about customer A's past purchase history."

[1001] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[1002] Example: AI system for medical institutions

[1003] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[1004] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[1005] The terminal works in conjunction with the hospital's systems and provides an interface for setting commonly used terms and medical procedures for doctors and nurses.

[1006] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, and the server generates an appropriate response and sends it to the terminal.

[1007] The server collects and analyzes the dialogue logs between users and the AI. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[1008] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[1009] In this way, the present invention provides an interactive artificial intelligence system optimized for a specific industry, which can greatly improve the business efficiency of an enterprise.

[1010] The processing flow will be explained below.

[1011] ---

[1012] Step 1:

[1013] The server collects data related to a particular industry from external databases and specialized books.

[1014] Step 2:

[1015] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[1016] Step 3:

[1017] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[1018] Step 4:

[1019] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[1020] Step 5:

[1021] Users input their company's specific business processes and terminology through the interface.

[1022] Step 6:

[1023] The server collects input data from users and reflects it in the AI ​​model.

[1024] Step 7:

[1025] Users use conversational artificial intelligence in their daily work, for example, by typing, "What's on the agenda for today's meetings?"

[1026] Step 8:

[1027] The terminal transmits the user's input to the server in real time.

[1028] Step 9:

[1029] The server generates a response based on the user's input and sends the response to the terminal.

[1030] Step 10:

[1031] The terminal presents the generated response to the user.

[1032] Step 11:

[1033] The server securely stores the dialogue logs between the user and the AI.

[1034] Step 12:

[1035] The server periodically retrieves the accumulated dialogue logs and performs data preprocessing.

[1036] Step 13:

[1037] The server analyzes the pre-processed data and identifies areas for improvement in the model.

[1038] Step 14:

[1039] The server retrains the conversational AI model based on the analysis results.

[1040] Step 15:

[1041] The server develops new industry-specific dialogue packages.

[1042] Step 16:

[1043] The terminal receives new interaction packages from the server and updates the system.

[1044] Step 17:

[1045] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[1046] Step 18:

[1047] The device sends the user's input to the server, and the conversational AI generates a response and returns it to the device.

[1048] Step 19:

[1049] The terminal presents the new response to the user.

[1050] Step 20:

[1051] The server continues to store updated interaction logs to help improve the accuracy of the system.

[1052] ---

[1053] These are the specific steps in programming an interactive AI system optimized for a specific industry.

[1054] Example 1

[1055] 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."

[1056] Conventional conversational AI systems have struggled to consistently provide everything from data collection to model training and customization, dialogue log accumulation and analysis, and automatic system updates to meet the specialized requirements of specific industries. As a result, they have been unable to provide highly accurate responses tailored to each company's needs. While this is expected to improve business efficiency, it actually places a heavy burden on users and lacks convenience. In particular, there is a need for appropriate and rapid response in the process of generating real-time responses, refining models, and updating dialogue packages based on feedback.

[1057] 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.

[1058] In this invention, the server includes a means for collecting data related to a specific industry, a means for training an interactive AI model using the collected data, and a means for generating responses to users. This makes it possible to collect data suited to the specialized needs of each company and generate highly accurate responses in real time. Furthermore, by supporting customization, analysis of dialogue logs, model retraining based on feedback, and automatic updates of dialogue packages, it is possible to improve business efficiency and reduce the burden on users.

[1059] "Means for collecting data related to a specific industry" refers to methods or devices for obtaining information related to a specific industry from external databases or specialized books.

[1060] "Means for training an interactive AI model using collected data" refers to a method or device for building an interactive AI model using a machine learning algorithm based on acquired data and optimizing the model.

[1061] "Means for customizing a company's business processes and terminology" refers to methods and devices that provide interfaces and configuration functions for reflecting the unique business procedures and terminology of each company within the system.

[1062] "Means for receiving input data from a user and transmitting it to a server in real time" refers to a communication protocol or device for instantly receiving data input by a user into the system and transmitting it to a server.

[1063] A "means for generating a response to a user" is a method or device that utilizes an interactive artificial intelligence model to generate an appropriate reply based on input data from a user.

[1064] "Means for analyzing dialogue logs and retraining models" refers to a method or device that analyzes dialogue history with a user and retrains the model based on this to improve the accuracy of the dialogue-based artificial intelligence model.

[1065] "Means for developing and updating new industry-specific dialogue packages" refers to methods and devices for creating new dialogue packages tailored to specific industries based on analysis results and feedback, and for keeping the system up to date.

[1066] "Means for delivering updated dialogue packages to terminals and automatically updating the system" refers to a method or device for applying the latest dialogue packages delivered from a server to terminals and automatically updating the system.

[1067] MODE FOR CARRYING OUT THE INVENTION

[1068] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[1069] server

[1070] The server provides the following means:

[1071] 1. A means of collecting data related to a specific industry: The server obtains information related to a specific industry from external databases or specialized books. For example, in the medical industry, data is collected from PubMed and other medical databases. During this process, the data is acquired and analyzed using Python's requests library, BeautifulSoup, and Scrapy.

[1072] 2. Means for training a conversational AI model using the collected data: The server uses the collected data to train a conversational AI model using TensorFlow or PyTorch. Specifically, it uses natural language processing (NLP) techniques to build a model that can appropriately respond to questions and requests specific to the industry.

[1073] 3. Means for generating a response to the user: The server uses an interactive artificial intelligence model to generate an appropriate response based on the input data from the user. For example, if a doctor inputs "Please tell me about patient X's allergies," the server will query the electronic medical record database and generate an appropriate response from the results.

[1074] 4. Analyzing the dialogue log and retraining the model: The server analyzes the dialogue history with the user and retrains the dialogue AI model to improve its accuracy. It uses Python's pandas and NumPy to analyze the log data and retrains the model based on the analysis results.

[1075] 5. Means for developing and updating new industry-specific dialogue packages: The server creates new dialogue packages tailored to specific industries based on analysis results and feedback, and updates the system to keep it up to date.

[1076] Terminal

[1077] The terminal provides the means to:

[1078] 1. A means to customize a company's business processes and terminology: The terminal provides an interface that reflects a company's specific business procedures and terminology within the system. For example, a customized interface can be built using a front-end framework such as React or Vue.js.

[1079] 2. Means for receiving input data from the user and transmitting it to the server in real time: The terminal uses a communication protocol (e.g., REST API or WebSocket) to transmit data input by the user to the server in real time.

[1080] 3. Means for delivering updated dialogue packages to terminals and automatically updating the system: The terminals receive the latest dialogue packages delivered from the server and automatically update the system, allowing users to always interact with the latest information.

[1081] User

[1082] The user provides the means to:

[1083] 1. Input means: A user uses an interactive AI system to input specific questions or commands into the system. For example, a specific request such as "Please tell me the allergy information of patient X" is input into a terminal.

[1084] 2. Feedback: The user provides feedback on the system's response. They evaluate whether the response is accurate or there is room for improvement, and send the feedback to the server via their device. This feedback is used to retrain the model and improve the system.

[1085] Example: AI system for medical institutions

[1086] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[1087] The server collects data on medical terminology and medical procedures from medical databases and uses TensorFlow to train an interactive AI model. For example, it can collect research paper data from PubMed via an API and use it to build a neural network model.

[1088] The device connects to the hospital's electronic medical record system and provides an interface for doctors and nurses to set commonly used terms and medical procedures. This is achieved through a web application written in React.

[1089] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, which retrieves the data from the electronic medical record, generates an appropriate response, and returns it to the terminal.

[1090] The server collects dialogue logs between users and the AI ​​and analyzes them using Python's pandas. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[1091] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[1092] Examples of prompt statements

[1093] Below are some example prompts from an AI model for healthcare:

[1094] "I'd like to see Patient X's most recent medical records. What procedures were performed?"

[1095] By inputting this prompt into a generative AI model, it is possible to extract appropriate medical information. Furthermore, it can also suggest improvements to the prompt based on the results of analyzing actual dialogue logs.

[1096] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1097] Step 1: Data Collection (Server)

[1098] Specific operation: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses Python's requests library to send an HTTP request to the PubMed API and parses the returned JSON response.

[1099] Input: Data related to a specific industry (e.g., PubMed API)

[1100] Data processing: HTTP request sending, response analysis

[1101] Output: Retrieved data (JSON format)

[1102] Step 2: Model training (server)

[1103] Specific operation: The server uses the collected data to train an interactive artificial intelligence model. For example, it uses the TensorFlow library to build a neural network model and supplies training data to train the model.

[1104] Input: Collected data (JSON format)

[1105] Data processing: Data preprocessing (cleaning, tokenization), model building, training

[1106] Output: A trained interactive AI model

[1107] Step 3: Interface Settings (Device)

[1108] Specific operation: The terminal provides an interface for inputting a company's business processes and terminology into the system. The terminal uses React to build a customized interface, allowing users to input company-specific information.

[1109] Input: Company business processes and terminology (user input)

[1110] Data processing: Interface construction

[1111] Output: Customized configuration information

[1112] Step 4: User Input (User)

[1113] Specific Actions: A user (e.g., a doctor or nurse) uses a conversational AI system to input specific questions or commands into the system, and transmits data to the system using text input or voice recognition technology (e.g., Google Speech-to-Text API).

[1114] Input: User questions or commands (text or voice)

[1115] Data processing: Voice recognition (when inputting voice)

[1116] Output: User input data (text format)

[1117] Step 5: Send data (terminal)

[1118] Specific operation: The terminal sends user input data to the server in real time. To do this, the terminal uses REST API or WebSocket to send data with low latency.

[1119] Input: User input data (text format)

[1120] Data processing: Data encoding, HTTP POST request or WebSocket transmission

[1121] Output: Data sent to the server

[1122] Step 6: Response Generation (Server)

[1123] Specific operation: The server generates a response using an interactive artificial intelligence model based on the received user input data. The server sends a query to the electronic medical record database and generates an appropriate response from the results.

[1124] Input: User input data (text format)

[1125] Data processing: query generation, database interrogation, response generation

[1126] Output: Generated response data (text format)

[1127] Step 7: Response presentation (terminal)

[1128] Specific operation: The terminal presents the response data returned from the server to the user. The terminal dynamically updates the web page and displays the response in a user-friendly format.

[1129] Input: Response data from the server (text format)

[1130] Data processing: Data decoding, display format conversion

[1131] Output: The response presented to the user

[1132] Step 8: Feedback (User)

[1133] Specific operation: The user provides feedback on the system's response by entering an evaluation and suggestions for improvement through the terminal and sending the feedback to the server.

[1134] Input: User feedback (text format)

[1135] Data processing: Evaluation data input

[1136] Output: Feedback sent to the server

[1137] Step 9: Log analysis and model update (server)

[1138] How it works: The server analyzes the collected interaction logs and feedback. It uses Python's pandas and NumPy to extract insights from the log data and retrain the model.

[1139] Input: Dialogue logs, feedback data

[1140] Data processing: log analysis, data preprocessing, model retraining

[1141] Output: Updated interactive artificial intelligence model

[1142] Step 10: Update packages (device)

[1143] Specific operation: The terminal receives new industry-specific dialogue packages distributed by the server and automatically updates the system, so that the latest dialogue models and functions are applied to the terminal.

[1144] Input: New dialogue package

[1145] Data processing: Package download, update application

[1146] Output: Updated system

[1147] (Application example 1)

[1148] 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."

[1149] In modern business, there is a growing demand for conversational AI systems to efficiently handle specific industries. In particular, the food delivery industry needs ways to streamline order taking, delivery tracking, and customer support. However, performing these tasks manually is time-consuming, labor-intensive, and prone to errors. Therefore, the present invention aims to provide an industry-specific conversational AI system to solve these problems.

[1150] 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.

[1151] In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive AI model using the collected data, means for customizing a company's business processes and terminology, means for receiving input data from a user and accumulating an interaction log, means for analyzing the accumulated interaction log and retraining the model, means for developing and updating a new industry-specific interaction package, means for providing the updated interaction package to the user, means for converting user voice input into text, means for analyzing user order information using the interactive AI model and providing menu options, and means for tracking delivery status for a specific order and providing the user with an estimated delivery time. This makes it possible to streamline a series of operations, from order taking to delivery tracking and customer support, even in the food delivery industry.

[1152] "Specific industry" refers to all businesses related to a specific business area or field.

[1153] "Data collection" refers to the process of obtaining information relevant to a particular industry.

[1154] "Conversational artificial intelligence model" refers to a machine learning model that is trained to respond appropriately to user input in natural language.

[1155] "A company's business processes" refers to the set of procedures and methods by which a company carries out its daily business operations.

[1156] "Jargon" refers to words and phrases specific to a particular industry or field.

[1157] "Customization" refers to the process of tailoring a system or service to meet specific needs and requirements.

[1158] "Input Data" refers to information or instructions provided by a user.

[1159] "Dialogue log" refers to a record of all conversations exchanged between a user and an interactive AI system.

[1160] "Data analysis" refers to the process of analyzing accumulated data and extracting meaningful information and patterns.

[1161] "Model retraining" refers to the process of retraining an existing machine learning model based on new data or analytical results.

[1162] "Dialogue package" refers to a software module that forms part of a dialogue-based artificial intelligence system optimized for a specific industry.

[1163] "Voice input" refers to the process of capturing user spoken words into the system.

[1164] "Menu options" refer to selectable products and services offered to a user.

[1165] "Delivery status" refers to the progress of an ordered item until it is delivered.

[1166] "Expected Delivery Time" means the estimated time it will take for an ordered item to be delivered.

[1167] The present invention provides an interactive artificial intelligence system optimized for a specific industry, streamlining order acceptance, delivery tracking, and customer support in the food delivery industry. This system is configured to operate in cooperation with each element: server, terminal, and user.

[1168] server

[1169] The server provides the following means:

[1170] 1. Data collection method: Collect information related to a specific industry from external databases to understand industry-specific terminology and business processes.

[1171] 2. Model training method: Based on the collected data, a conversational artificial intelligence model (generative AI model) is trained, which builds a model that can respond appropriately to industry-specific questions and requests.

[1172] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[1173] 4. Package development tools: Based on the analysis results, we will develop new industry-specific interactive packages to optimize order taking, delivery tracking, and customer support in the food delivery industry.

[1174] Terminal

[1175] The terminal provides the means to:

[1176] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[1177] 2. Data reception and transmission means: Data input from the user (e.g., order details or questions) is transmitted to the server in real time, and response data from the server is presented to the user.

[1178] 3. Voice input means: Provides a voice recognition function to convert voice input into text.

[1179] 4. Package update means: Receive new industry-specific dialogue packages (generative AI models) distributed from the server and update the system.

[1180] User

[1181] The user provides the means to:

[1182] 1. Input method: Food delivery orders and inquiries are input into the conversational AI system. For example, you can say, "I'd like to order one Margherita pizza and one Coke Zero" by voice.

[1183] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[1184] Hardware and software used

[1185] Hardware: Cloud servers (AWS, Google Cloud, etc.), smartphones (iOS, Android, etc.)

[1186] Software: Generative AI model (facebook / blenderbot-400M-distill), framework (Flask), speech recognition API (Google Cloud Speech-to-Text), database (MongoDB)

[1187] Data processing and calculation

[1188] Speech-to-text conversion: Uses the Google Cloud Speech-to-Text API to convert speech to text, allowing you to capture user speech as text.

[1189] Dialogue model response generation: A generative AI model trained on collected data is used to generate appropriate responses to user input. For example, if a user inputs, "I'd like to order one Margherita pizza and one Coke Zero," the conversational AI system will provide corresponding menu options.

[1190] Delivery status tracking and provision: Real-time delivery status information for specific orders is tracked and provided to users. For example, if a user requests "What is the delivery status of order number 12345?", the current delivery status and estimated delivery time will be provided.

[1191] Specific examples

[1192] Order Taker: "I'd like to order one Margherita pizza and one Coke Zero."

[1193] Delivery tracking: "What's the delivery status of order number 12345?"

[1194] Customer Support: "Why can't I use my coupon code?"

[1195] Prompt Sentence Examples

[1196] Enter your order: "I'd like to order one Margherita pizza and one Coke Zero."

[1197] Delivery tracking: "What's the delivery status of order number 12345?"

[1198] Support Enquiry: "Why can't I use my coupon code?"

[1199] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1200] Step 1:

[1201] Users voice-order food delivery through a smartphone application, where the voice input is captured.

[1202] Input: User speaks "I would like to order one Margherita pizza and one Coke Zero."

[1203] Output: Audio input data

[1204] Step 2:

[1205] The device converts voice input into text using the Google Cloud Speech-to-Text API.

[1206] Input: Voice input data

[1207] Output: Text data "I would like to order one Margherita pizza and one Coke Zero."

[1208] Specific operation: Call the voice recognition API and obtain the voice data as text data.

[1209] Step 3:

[1210] The device sends text data to the server, which uses a generative AI model (facebook / blenderbot-400M-distill) to generate an appropriate response to the user's order.

[1211] Input: Text data "I would like to order one Margherita pizza and one Coke Zero."

[1212] Output: Response data containing menu options

[1213] Specific operation: Text data is input into an interactive artificial intelligence model and response text is generated.

[1214] Step 4:

[1215] The server transmits the generated response data to the terminal, which then presents it to the user.

[1216] Input: Response data "Would you like one Margherita pizza and one Coke Zero?"

[1217] Output: The response text to display to the user

[1218] Specific behavior: Display the response text on the device's user interface.

[1219] Step 5:

[1220] The user provides confirmation, for example, a voice or text reply of "Yes, please."

[1221] Input: User response "Yes, please"

[1222] Output: Input data for verification

[1223] Step 6:

[1224] The terminal performs voice recognition again and sends the result as text data to the server. The server receives this confirmation data and processes it to confirm the order.

[1225] Input: Input data for confirmation "Yes, please"

[1226] Output: Order confirmation data

[1227] Specific operation: Converts speech into text and sends the data to the server to confirm the order. The server stores the order information in a database.

[1228] Step 7:

[1229] After the order is confirmed, the server generates a tracking ID for tracking the delivery status and sends it to the terminal. The terminal displays the tracking ID to the user.

[1230] Input: Order confirmation data

[1231] Output: Tracking ID

[1232] Specific operation: A tracking ID is generated based on the order information and sent to the terminal, which then displays the tracking ID on the user interface.

[1233] Step 8:

[1234] A user queries the application to check the delivery status of an order, entering something like "What is the delivery status of order number 12345?"

[1235] Input: User inquiry: "What is the delivery status of order number 12345?"

[1236] Output: Delivery tracking request data

[1237] Step 9:

[1238] The server tracks the delivery status, obtains the status in real time, and transmits the data to the terminal, which then presents the status data to the user.

[1239] Input: Delivery tracking request data

[1240] Output: Delivery status data "Current delivery status is on its way and will arrive in 15 minutes."

[1241] Specific operation: Obtain the current status from the delivery tracking system and send the information to the terminal.

[1242] Step 10:

[1243] The order is completed when the user receives it. The user provides feedback to the system, for example, by typing "The item arrived safely."

[1244] Input: User feedback "The item arrived safely."

[1245] Output: Feedback data

[1246] Specific operation: Feedback data is sent to the server and used to improve the system.

[1247] 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.

[1248] ---

[1249] This invention is a system that provides more sophisticated responses by combining a conversational AI system optimized for a specific industry with an emotion engine that recognizes the user's emotions. This system is configured so that the server, terminal, and user elements work together and the emotion engine is used to realize responses based on the user's emotional state.

[1250] server

[1251] The server provides the following means:

[1252] 1. Data collection methods: Data related to a specific industry are collected from external databases and specialized books.

[1253] 2. Model training method: Train an interactive artificial intelligence model based on the collected data.

[1254] 3. Emotion engine: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response.

[1255] 4. Data analysis: Analyze the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights.

[1256] 5. Package development method: Based on the analysis results, new industry-specific dialogue packages will be developed.

[1257] Terminal

[1258] The terminal provides the means to:

[1259] 1. Customization methods: Provide an interface for inputting a company's business processes and terminology into the system.

[1260] 2. Data reception and transmission means: Input data and emotion data from the user are transmitted to the server in real time, and response data from the server is presented to the user.

[1261] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[1262] User

[1263] The user provides the means to:

[1264] 1. Input means: Use conversational artificial intelligence in your daily work by inputting questions and commands into the system.

[1265] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This data is sent to the emotion engine.

[1266] 3. Feedback measures: Providing feedback on the system's response.

[1267] Example: Emotion recognition AI system for medical institutions

[1268] As a specific example, an embodiment of an emotion-recognition conversational artificial intelligence system for medical institutions will be described below:

[1269] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[1270] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[1271] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[1272] Users (doctors and nurses) use conversational AI in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergies," the device sends this input data and the patient's emotional data to the server.

[1273] The server generates an appropriate and emotionally sensitive response based on the user input and emotion data and sends it to the terminal.

[1274] The terminal presents the generated response to the user.

[1275] The server collects and analyzes user-AI dialogue logs and emotional data, and uses the results of this analysis to retrain the model and develop a new dialogue package specialized for medical work.

[1276] In this way, by combining a conversational AI system optimized for a specific industry with an emotion engine, the present invention can provide more precise and attentive responses to users, significantly improving business efficiency for companies.

[1277] The processing flow will be explained below.

[1278] ---

[1279] Step 1:

[1280] The server collects data related to a particular industry from external databases and specialized books.

[1281] Step 2:

[1282] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[1283] Step 3:

[1284] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[1285] Step 4:

[1286] The server incorporates an emotion engine that recognizes the user's emotions, allowing it to analyze the emotion data and reflect it in responses.

[1287] Step 5:

[1288] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[1289] Step 6:

[1290] Users input their company's specific business processes and terminology through the interface.

[1291] Step 7:

[1292] The server collects input data from users and reflects it in the AI ​​model.

[1293] Step 8:

[1294] Users use conversational AI in their daily work, utilizing cameras and microphones to capture emotional data, for example by typing, "What's on the agenda for today's meeting?"

[1295] Step 9:

[1296] The terminal transmits the user's input data and emotion data to the server in real time.

[1297] Step 10:

[1298] The server generates a response based on the user's input and emotion data and sends the response to the terminal.

[1299] Step 11:

[1300] The terminal presents the generated response to the user.

[1301] Step 12:

[1302] The server securely stores the dialogue logs and emotional data between the user and the AI.

[1303] Step 13:

[1304] The server periodically retrieves the accumulated dialogue logs and emotion data and performs data preprocessing.

[1305] Step 14:

[1306] The server analyzes the preprocessed data and sentiment data to identify areas for improvement in the model.

[1307] Step 15:

[1308] The server retrains the conversational AI model based on the analysis results.

[1309] Step 16:

[1310] The server develops new industry-specific dialogue packages.

[1311] Step 17:

[1312] The terminal receives new interaction packages from the server and updates the system.

[1313] Step 18:

[1314] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[1315] Step 19:

[1316] The device sends the user's input data and emotional data to the server, and the conversational AI generates an appropriate response and returns it to the device.

[1317] Step 20:

[1318] The terminal presents the new response to the user.

[1319] Step 21:

[1320] The server continues to store updated dialogue logs and emotional data to help improve the accuracy of the system.

[1321] ---

[1322] These are the specific steps in programming a conversational AI system that combines an emotion engine and is optimized for a specific industry.

[1323] Example 2

[1324] 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."

[1325] Conventional conversational AI systems generate responses only in response to user input and do not take into account the user's emotional state, which can result in a lack of appropriate responses and consideration. Furthermore, generating responses based on the user's emotional state is difficult, which reduces the accuracy and reliability of systems specialized for specific industries.

[1326] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for recognizing and acquiring user emotion data, means for receiving input data and emotion data from the user and analyzing them in real time, means for generating an appropriate response based on the analyzed data, means for presenting the generated response to the user, means for analyzing accumulated dialogue logs and emotion data and retraining the model, means for developing and updating a new industry-specific dialogue package, and means for providing the updated dialogue package to the user. This makes it possible to generate responses according to the user's emotional state, thereby significantly improving business efficiency and reliability in a specific industry.

[1327] "Means for collecting data related to a specific industry" refers to systems or software for automatically collecting information and data related to a specific industry from external databases or specialized books.

[1328] A "means for training an interactive artificial intelligence model" is a method for training an interactive artificial intelligence model using a machine learning algorithm based on collected data.

[1329] "Means for recognizing and acquiring user emotional data" refers to technology that recognizes a user's facial expressions and tone of voice through devices such as a camera or microphone, and acquires their emotional state as data.

[1330] "Means for receiving input data and emotional data from a user and analyzing it in real time" refers to a system in which text, voice data, and emotional data sent by a user are instantly received by a server and analyzed.

[1331] The "means for generating an appropriate response based on the analyzed data" is a method for generating an appropriate response based on user input and emotion data analyzed in real time.

[1332] The "means for presenting the generated response to the user" is a system that displays or audibly conveys the response generated by the server to the user via the terminal.

[1333] "Means for analyzing accumulated dialogue logs and emotional data and retraining a model" refers to a method for analyzing dialogue history and emotional data, and retraining an interactive artificial intelligence model based on the analysis to improve its performance.

[1334] "Means for developing and updating new industry-specific dialogue packages" refers to technology that creates dialogue scenarios and response models optimized for specific industries based on analysis results and new data, and updates the system.

[1335] The "means for providing an updated dialogue package to a user" refers to a method for delivering a newly developed or updated dialogue artificial intelligence package to a user's terminal and making it available for use.

[1336] This invention is a system that provides more accurate responses by combining an emotional engine with a conversational AI system optimized for a specific industry. This system works in conjunction with the server, terminal, and user elements to realize responses based on the user's emotional state.

[1337] Server Features

[1338] The server provides the following main functions:

[1339] 1. Data collection method: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses the PubMed API to obtain medical-related data.

[1340] 2. Model training method: Train an interactive AI model based on the collected data. This method uses machine learning frameworks such as TensorFlow and PyTorch. For example, build an interactive model of the medical treatment process based on medical data.

[1341] 3. Emotion Engine: The server uses emotion recognition software such as Emotion API to analyze the user's facial expressions and tone of voice and obtain emotional data, which is then reflected in the generation of a response.

[1342] 4. Data analysis: Analyzes the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights, thereby improving the efficiency and accuracy of the entire system.

[1343] 5. Package development: Based on the analysis results, new industry-specific dialogue packages are developed and the system is updated. This method is important for generating new dialogue scenarios and response models.

[1344] Device Features

[1345] The terminal provides the following main features:

[1346] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system. For example, in a hospital system, it sets the terminology commonly used by doctors and nurses.

[1347] 2. Data reception and transmission means: Sends input data and emotion data from the user to the server in real time, and presents the response from the server to the user. For this function, for example, the Zoom SDK can be used.

[1348] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system, ensuring that the latest data and response models are always available.

[1349] User operations

[1350] The user does the following:

[1351] 1. Input Method: Conversational AI is used in daily operations to input questions or commands into the system. For example, a doctor might type, "Please tell me Patient X's allergy information." This input can be done via keyboard, voice, or touch panel.

[1352] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This acquired emotion data is sent to the emotion engine on the server.

[1353] 3. Feedback measures: Providing feedback on the system's response. This feedback helps refine and improve the system.

[1354] Specific examples

[1355] For example, an emotion-aware conversational AI system for a medical institution:

[1356] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[1357] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[1358] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[1359] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice.

[1360] Prompt Sentence Examples

[1361] "Please tell me Patient X's allergy information and current emotional state."

[1362] "Please advise on a treatment process that takes into consideration the patient's feelings."

[1363] "Please provide examples of responses for specific medical conditions."

[1364] This invention combines an interactive artificial intelligence system optimized for a specific industry with an emotion engine, which is expected to realize highly accurate responses that are attuned to the user, significantly improving the efficiency and reliability of business operations.

[1365] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1366] Step 1: Data collection

[1367] The server collects data related to a specific industry from external databases and specialized books. The input for this process is a specific keyword or query, and the output is related data. For example, in a system for medical institutions, the server uses the PubMed API to automatically collect literature data based on keywords such as "allergy" and "medical treatment process." The collected data is then used in subsequent model training.

[1368] Step 2: Model training

[1369] The server trains an interactive AI model based on the collected data. The input for this process is the collected data, and the output is a trained interactive AI model. Specifically, it uses machine learning frameworks such as TensorFlow and PyTorch to build and train dialogue scenarios based on medical treatment processes using medical data. The trained model is then used to generate responses to questions from users.

[1370] Step 3: User Input

[1371] The user inputs a question or command into the system. The input for this process is text or voice input from the user, and the output is the input data. For example, a doctor might input "Please tell me about patient X's allergy information" into a terminal. This input data is used in the subsequent emotion recognition and response generation steps.

[1372] Step 4: Emotion Recognition

[1373] The device acquires emotion data from the user's facial expressions and tone of voice via a camera and microphone. The input for this process is the user's visual and audio characteristics, and the output is emotion data. Specifically, OpenCV is used to analyze the doctor's facial expressions captured by the camera to determine whether the user is feeling stressed. The emotion data is sent to the server and used in the subsequent response generation step.

[1374] Step 5: Response Generation

[1375] The server generates an appropriate response based on the user's input data and emotional data. The input for this process is the user's question or command, as well as emotional data, and the output is a response based on that. Specifically, using a generative AI model (e.g., GPT-4), if a user inputs "Tell me about patient X's allergies," the server generates a response based on the emotional data, such as "Patient X's allergy information is ____. Also, a high stress level has been detected, so special care is required."

[1376] Step 6: Present the response

[1377] The terminal presents the generated response to the user. The input for this process is the response data received from the server, and the output is the content presented to the user. Specifically, the terminal's display and voice output function are used to notify the doctor, "Patient X's allergy information is ____. Also, it appears that his current stress level is high, so he requires special attention."

[1378] Step 7: Data analysis and model updating

[1379] The server collects and analyzes user-AI dialogue logs and emotional data. The inputs for this process are dialogue logs and emotional data, and the output is the analysis results and an improved dialogue model. Specifically, the accuracy of the model is improved and new insights are extracted based on the dialogue logs and emotional data, and the model is retrained using TensorFlow. The analysis results are used to develop new industry-specific dialogue packages and update the system.

[1380] (Application example 2)

[1381] 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."

[1382] Modern brick-and-mortar stores require effective methods to improve customer shopping experiences. However, conventional methods struggle to grasp customers' emotional states in real time and provide appropriate product recommendations. Furthermore, conversational AI systems generate responses without considering customers' emotions, resulting in low customer satisfaction. Therefore, it is important to provide personalized product recommendations based on customers' emotional states.

[1383] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for customizing the company's business processes and terminology, emotion engine means for recognizing the user's emotional state, analyzing the emotion data, and reflecting the emotion data in responses, and data processing means for providing product suggestions based on the customer's emotional state. This makes it possible to provide personalized product suggestions based on the customer's emotional state in real time.

[1384] "Specific industry" refers to a specific commercial or business sector that specializes in handling data and business processes related to that industry.

[1385] "Data collection means" refers to the means of obtaining information and data related to a specific industry from external databases and materials.

[1386] An "interactive artificial intelligence model" refers to an artificial intelligence algorithm or program that enables natural dialogue with the user.

[1387] A "model training means" is a means for training an interactive artificial intelligence model using collected data to improve its performance.

[1388] "Business process" refers to a series of procedures and flows that a company follows when carrying out its business.

[1389] "Jargon" refers to specialized words and phrases used in a particular industry or field.

[1390] "Customization tools" refers to configuration tools and interfaces that allow a company to adapt the system to its specific business processes and terminology.

[1391] "Input data" refers to information and instructions provided by a user to a system.

[1392] "Interaction log" refers to a record of interaction data between a user and a system.

[1393] "Analytical tools" refers to techniques and methods for analyzing accumulated data and using the results to improve the model.

[1394] "Emotion engine means" refers to technology or algorithms for recognizing the user's emotional state, analyzing that data, and reflecting it in dialogue responses.

[1395] "Data processing means" refers to a mechanism for analyzing acquired data and performing specific processing.

[1396] "Camera and microphone means" refers to a photographic and audio recording device for capturing the customer's facial expressions and tone of voice in real time.

[1397] "Real-time" means near-simultaneous data processing and response.

[1398] "Personalized product proposals" refer to proposals for products and services that are individually customized based on the customer's individual attributes and circumstances.

[1399] The present invention is a system that provides personalized product recommendations using a conversational AI system combined with an emotion engine that recognizes user emotions. This system collects data related to a specific industry, trains a conversational AI model, and customizes the company's business processes and terminology to provide optimized services to customers.

[1400] server

[1401] The server provides the following means:

[1402] 1. Data collection methods: Collect data related to a specific industry from external databases and sources. For example, in the distribution industry, collect and use consumer behavior data.

[1403] 2. Model training method: The collected data is used to train a conversational artificial intelligence model. The model used here is a generative AI model that learns from a large amount of conversational data.

[1404] 3. Emotion engine means: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response. This engine uses the DeepFace library to recognize emotions from facial expressions and tone of voice.

[1405] 4. Data processing means: Processes data to provide product suggestions based on the customer's emotional state. This means obtains product recommendations using external APIs.

[1406] 5. Analysis method: Preprocess the accumulated dialogue logs and perform data analysis, which allows us to retrain the model and improve its accuracy.

[1407] Terminal

[1408] The terminal provides the means to:

[1409] 1. Customization: Provides an interface for setting up a company's business processes and terminology. For example, inputting terminology specific to the distribution industry.

[1410] 2. Input data receiving means: Provides a communication means for transmitting input data from the user to the server in real time.

[1411] 3. Emotion recognition means: Equipped with a camera and microphone to capture the customer's facial expressions and tone of voice in real time. For example, the camera and microphone of a smartphone can be used to recognize the customer's emotions.

[1412] 4. Response providing means: Provides the product suggestions sent from the server to the user. For example, displays the acquired product data on the smartphone screen.

[1413] User

[1414] The user provides the means to:

[1415] 1. Input method: Customers input their questions or opinions using conversational AI. For example, they can input "What's your recommendation today?" through a smartphone app.

[1416] 2. Emotion recognition: Emotion data is acquired from the user's facial expressions and tone of voice via a camera and microphone. This data is sent to the server in real time.

[1417] 3. Feedback measures: Providing feedback on the system's response, e.g., rating satisfaction with a suggested product.

[1418] Specific examples

[1419] For example, if the user is feeling stressed, the system will suggest relaxation-related products, and if the user is smiling, entertainment products that will increase their enjoyment.

[1420] Prompt Sentence Examples

[1421] "The emotion the customer is feeling is stress. Make product recommendations based on this emotion."

[1422] In this way, by using an interactive artificial intelligence system combined with an emotion engine, the present invention can provide personalized services based on the customer's emotional state in real time, thereby improving the shopping experience in physical stores.

[1423] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1424] Step 1:

[1425] The device captures the customer's facial expressions and tone of voice using a camera and microphone. The input is the customer's facial expression data and voice data. This data is collected in real time and saved as image and audio files.

[1426] Step 2:

[1427] The device sends the collected facial expression data and voice data to the server. The input is the image and voice files collected in step 1. Uploading this to the server enables emotion analysis.

[1428] Step 3:

[1429] The server analyzes the facial expression data using the DeepFace library to identify the customer's emotional state. The input is the image file sent in step 2. This data is passed to DeepFace for processing and the output is the emotion analysis result.

[1430] Step 4:

[1431] The server generates a prompt to suggest products appropriate for the customer based on their emotional state and sends a request to the external API. The input is the emotion analysis result obtained in step 3. Based on this, the server generates a prompt such as "The customer is feeling ____. Please suggest products based on this emotion," and sends it to the external API.

[1432] Step 5:

[1433] An external API generates product suggestions based on the prompt sent. The input is the prompt sent in step 4. The API parses it and generates a list of relevant products. The output is the data of the recommended products.

[1434] Step 6:

[1435] The server receives the product suggestions obtained from the external API. The input is the product list data generated in step 5. This data is stored in the server and prepared for transmission to the device.

[1436] Step 7:

[1437] The terminal displays the product proposals received from the server to the customer. The input is the product list data received in step 6. This data is displayed on the smartphone screen, providing specific product information to be proposed to the customer.

[1438] Step 8:

[1439] The user provides feedback on the proposed product. The input is the product information displayed in step 7. The user enters their satisfaction level and opinions about the product and sends them to the server via their terminal.

[1440] Step 9:

[1441] The server analyzes the collected feedback data and uses it to retrain the AI ​​model. The input is the feedback data obtained in step 8. The output is a new and improved AI model.

[1442] In this way, a system is completed that provides personalized product suggestions based on the user's emotional state.

[1443] 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.

[1444] 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.

[1445] 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.

[1446] [Fourth embodiment]

[1447] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1448] 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.

[1449] 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).

[1450] 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.

[1451] 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.

[1452] 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).

[1453] 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.

[1454] 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.

[1455] 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.

[1456] 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.

[1457] 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.

[1458] 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.

[1459] 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."

[1460] ---

[1461] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[1462] server

[1463] The server provides the following means:

[1464] 1. Data collection method: Collect data related to a specific industry from external databases and specialized books, thereby understanding industry-specific terminology and business processes.

[1465] 2. Model training method: Based on the collected data, an interactive artificial intelligence model is trained, which is built to respond appropriately to industry-specific questions and requests.

[1466] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[1467] 4. Package development method: Based on the analysis results, we develop new industry-specific dialogue packages, thereby providing customized solutions for each company.

[1468] Terminal

[1469] The terminal provides the means to:

[1470] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[1471] 2. Data reception and transmission means: Sends input data from the user to the server in real time, and presents response data from the server to the user.

[1472] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[1473] User

[1474] The user provides the means to:

[1475] 1. Input: Using conversational AI in everyday tasks, you input questions and commands into the system, such as specific requests like "What's on the agenda for today's meeting?" or "Tell me about customer A's past purchase history."

[1476] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[1477] Example: AI system for medical institutions

[1478] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[1479] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[1480] The terminal works in conjunction with the hospital's systems and provides an interface for setting commonly used terms and medical procedures for doctors and nurses.

[1481] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, and the server generates an appropriate response and sends it to the terminal.

[1482] The server collects and analyzes the dialogue logs between users and the AI. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[1483] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[1484] In this way, the present invention provides an interactive artificial intelligence system optimized for a specific industry, which can greatly improve the business efficiency of an enterprise.

[1485] The processing flow will be explained below.

[1486] ---

[1487] Step 1:

[1488] The server collects data related to a particular industry from external databases and specialized books.

[1489] Step 2:

[1490] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[1491] Step 3:

[1492] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[1493] Step 4:

[1494] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[1495] Step 5:

[1496] Users input their company's specific business processes and terminology through the interface.

[1497] Step 6:

[1498] The server collects input data from users and reflects it in the AI ​​model.

[1499] Step 7:

[1500] Users use conversational artificial intelligence in their daily work, for example, by typing, "What's on the agenda for today's meetings?"

[1501] Step 8:

[1502] The terminal transmits the user's input to the server in real time.

[1503] Step 9:

[1504] The server generates a response based on the user's input and sends the response to the terminal.

[1505] Step 10:

[1506] The terminal presents the generated response to the user.

[1507] Step 11:

[1508] The server securely stores the dialogue logs between the user and the AI.

[1509] Step 12:

[1510] The server periodically retrieves the accumulated dialogue logs and performs data preprocessing.

[1511] Step 13:

[1512] The server analyzes the pre-processed data and identifies areas for improvement in the model.

[1513] Step 14:

[1514] The server retrains the conversational AI model based on the analysis results.

[1515] Step 15:

[1516] The server develops new industry-specific dialogue packages.

[1517] Step 16:

[1518] The terminal receives new interaction packages from the server and updates the system.

[1519] Step 17:

[1520] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[1521] Step 18:

[1522] The device sends the user's input to the server, and the conversational AI generates a response and returns it to the device.

[1523] Step 19:

[1524] The terminal presents the new response to the user.

[1525] Step 20:

[1526] The server continues to store updated interaction logs to help improve the accuracy of the system.

[1527] ---

[1528] These are the specific steps in programming an interactive AI system optimized for a specific industry.

[1529] Example 1

[1530] 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."

[1531] Conventional conversational AI systems have struggled to consistently provide everything from data collection to model training and customization, dialogue log accumulation and analysis, and automatic system updates to meet the specialized requirements of specific industries. As a result, they have been unable to provide highly accurate responses tailored to each company's needs. While this is expected to improve business efficiency, it actually places a heavy burden on users and lacks convenience. In particular, there is a need for appropriate and rapid response in the process of generating real-time responses, refining models, and updating dialogue packages based on feedback.

[1532] 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.

[1533] In this invention, the server includes a means for collecting data related to a specific industry, a means for training an interactive AI model using the collected data, and a means for generating responses to users. This makes it possible to collect data suited to the specialized needs of each company and generate highly accurate responses in real time. Furthermore, by supporting customization, analysis of dialogue logs, model retraining based on feedback, and automatic updates of dialogue packages, it is possible to improve business efficiency and reduce the burden on users.

[1534] "Means for collecting data related to a specific industry" refers to methods or devices for obtaining information related to a specific industry from external databases or specialized books.

[1535] "Means for training an interactive AI model using collected data" refers to a method or device for building an interactive AI model using a machine learning algorithm based on acquired data and optimizing the model.

[1536] "Means for customizing a company's business processes and terminology" refers to methods and devices that provide interfaces and configuration functions for reflecting the unique business procedures and terminology of each company within the system.

[1537] "Means for receiving input data from a user and transmitting it to a server in real time" refers to a communication protocol or device for instantly receiving data input by a user into the system and transmitting it to a server.

[1538] A "means for generating a response to a user" is a method or device that utilizes an interactive artificial intelligence model to generate an appropriate reply based on input data from a user.

[1539] "Means for analyzing dialogue logs and retraining models" refers to a method or device that analyzes dialogue history with a user and retrains the model based on this to improve the accuracy of the dialogue-based artificial intelligence model.

[1540] "Means for developing and updating new industry-specific dialogue packages" refers to methods and devices for creating new dialogue packages tailored to specific industries based on analysis results and feedback, and for keeping the system up to date.

[1541] "Means for delivering updated dialogue packages to terminals and automatically updating the system" refers to a method or device for applying the latest dialogue packages delivered from a server to terminals and automatically updating the system.

[1542] MODE FOR CARRYING OUT THE INVENTION

[1543] This invention provides a dialogue-based AI system optimized for a specific industry, accumulates data through use in the daily operations of companies, and develops and provides specialized dialogue packages. This system is configured so that each element, including the server, terminal, and user, works in cooperation with each other.

[1544] server

[1545] The server provides the following means:

[1546] 1. A means of collecting data related to a specific industry: The server obtains information related to a specific industry from external databases or specialized books. For example, in the medical industry, data is collected from PubMed and other medical databases. During this process, the data is acquired and analyzed using Python's requests library, BeautifulSoup, and Scrapy.

[1547] 2. Means for training a conversational AI model using the collected data: The server uses the collected data to train a conversational AI model using TensorFlow or PyTorch. Specifically, it uses natural language processing (NLP) techniques to build a model that can appropriately respond to questions and requests specific to the industry.

[1548] 3. Means for generating a response to the user: The server uses an interactive artificial intelligence model to generate an appropriate response based on the input data from the user. For example, if a doctor inputs "Please tell me about patient X's allergies," the server will query the electronic medical record database and generate an appropriate response from the results.

[1549] 4. Analyzing the dialogue log and retraining the model: The server analyzes the dialogue history with the user and retrains the dialogue AI model to improve its accuracy. It uses Python's pandas and NumPy to analyze the log data and retrains the model based on the analysis results.

[1550] 5. Means for developing and updating new industry-specific dialogue packages: The server creates new dialogue packages tailored to specific industries based on analysis results and feedback, and updates the system to keep it up to date.

[1551] Terminal

[1552] The terminal provides the means to:

[1553] 1. A means to customize a company's business processes and terminology: The terminal provides an interface that reflects a company's specific business procedures and terminology within the system. For example, a customized interface can be built using a front-end framework such as React or Vue.js.

[1554] 2. Means for receiving input data from the user and transmitting it to the server in real time: The terminal uses a communication protocol (e.g., REST API or WebSocket) to transmit data input by the user to the server in real time.

[1555] 3. Means for delivering updated dialogue packages to terminals and automatically updating the system: The terminals receive the latest dialogue packages delivered from the server and automatically update the system, allowing users to always interact with the latest information.

[1556] User

[1557] The user provides the means to:

[1558] 1. Input means: A user uses an interactive AI system to input specific questions or commands into the system. For example, a specific request such as "Please tell me the allergy information of patient X" is input into a terminal.

[1559] 2. Feedback: The user provides feedback on the system's response. They evaluate whether the response is accurate or there is room for improvement, and send the feedback to the server via their device. This feedback is used to retrain the model and improve the system.

[1560] Example: AI system for medical institutions

[1561] As a specific example, an embodiment of an interactive artificial intelligence system for a medical institution will be described below:

[1562] The server collects data on medical terminology and medical procedures from medical databases and uses TensorFlow to train an interactive AI model. For example, it can collect research paper data from PubMed via an API and use it to build a neural network model.

[1563] The device connects to the hospital's electronic medical record system and provides an interface for doctors and nurses to set commonly used terms and medical procedures. This is achieved through a web application written in React.

[1564] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergy information," the terminal sends this input data to the server, which retrieves the data from the electronic medical record, generates an appropriate response, and returns it to the terminal.

[1565] The server collects dialogue logs between users and the AI ​​and analyzes them using Python's pandas. The results of this analysis are used to retrain the model and develop a new dialogue package specialized for medical work.

[1566] When new dialogue packages are developed, the terminal receives them and updates the system, allowing users to receive more specialized and precise medical support.

[1567] Examples of prompt statements

[1568] Below are some example prompts from an AI model for healthcare:

[1569] "I'd like to see Patient X's most recent medical records. What procedures were performed?"

[1570] By inputting this prompt into a generative AI model, it is possible to extract appropriate medical information. Furthermore, it can also suggest improvements to the prompt based on the results of analyzing actual dialogue logs.

[1571] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1572] Step 1: Data Collection (Server)

[1573] Specific operation: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses Python's requests library to send an HTTP request to the PubMed API and parses the returned JSON response.

[1574] Input: Data related to a specific industry (e.g., PubMed API)

[1575] Data processing: HTTP request sending, response analysis

[1576] Output: Retrieved data (JSON format)

[1577] Step 2: Model training (server)

[1578] Specific operation: The server uses the collected data to train an interactive artificial intelligence model. For example, it uses the TensorFlow library to build a neural network model and supplies training data to train the model.

[1579] Input: Collected data (JSON format)

[1580] Data processing: Data preprocessing (cleaning, tokenization), model building, training

[1581] Output: A trained interactive AI model

[1582] Step 3: Interface Settings (Device)

[1583] Specific operation: The terminal provides an interface for inputting a company's business processes and terminology into the system. The terminal uses React to build a customized interface, allowing users to input company-specific information.

[1584] Input: Company business processes and terminology (user input)

[1585] Data processing: Interface construction

[1586] Output: Customized configuration information

[1587] Step 4: User Input (User)

[1588] Specific Actions: A user (e.g., a doctor or nurse) uses a conversational AI system to input specific questions or commands into the system, and transmits data to the system using text input or voice recognition technology (e.g., Google Speech-to-Text API).

[1589] Input: User questions or commands (text or voice)

[1590] Data processing: Voice recognition (when inputting voice)

[1591] Output: User input data (text format)

[1592] Step 5: Send data (terminal)

[1593] Specific operation: The terminal sends user input data to the server in real time. To do this, the terminal uses REST API or WebSocket to send data with low latency.

[1594] Input: User input data (text format)

[1595] Data processing: Data encoding, HTTP POST request or WebSocket transmission

[1596] Output: Data sent to the server

[1597] Step 6: Response Generation (Server)

[1598] Specific operation: The server generates a response using an interactive artificial intelligence model based on the received user input data. The server sends a query to the electronic medical record database and generates an appropriate response from the results.

[1599] Input: User input data (text format)

[1600] Data processing: query generation, database interrogation, response generation

[1601] Output: Generated response data (text format)

[1602] Step 7: Response presentation (terminal)

[1603] Specific operation: The terminal presents the response data returned from the server to the user. The terminal dynamically updates the web page and displays the response in a user-friendly format.

[1604] Input: Response data from the server (text format)

[1605] Data processing: Data decoding, display format conversion

[1606] Output: The response presented to the user

[1607] Step 8: Feedback (User)

[1608] Specific operation: The user provides feedback on the system's response by entering an evaluation and suggestions for improvement through the terminal and sending the feedback to the server.

[1609] Input: User feedback (text format)

[1610] Data processing: Evaluation data input

[1611] Output: Feedback sent to the server

[1612] Step 9: Log analysis and model update (server)

[1613] How it works: The server analyzes the collected interaction logs and feedback. It uses Python's pandas and NumPy to extract insights from the log data and retrain the model.

[1614] Input: Dialogue logs, feedback data

[1615] Data processing: log analysis, data preprocessing, model retraining

[1616] Output: Updated interactive artificial intelligence model

[1617] Step 10: Update packages (device)

[1618] Specific operation: The terminal receives new industry-specific dialogue packages distributed by the server and automatically updates the system, so that the latest dialogue models and functions are applied to the terminal.

[1619] Input: New dialogue package

[1620] Data processing: Package download, update application

[1621] Output: Updated system

[1622] (Application example 1)

[1623] 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."

[1624] In modern business, there is a growing demand for conversational AI systems to efficiently handle specific industries. In particular, the food delivery industry needs ways to streamline order taking, delivery tracking, and customer support. However, performing these tasks manually is time-consuming, labor-intensive, and prone to errors. Therefore, the present invention aims to provide an industry-specific conversational AI system to solve these problems.

[1625] 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.

[1626] In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive AI model using the collected data, means for customizing a company's business processes and terminology, means for receiving input data from a user and accumulating an interaction log, means for analyzing the accumulated interaction log and retraining the model, means for developing and updating a new industry-specific interaction package, means for providing the updated interaction package to the user, means for converting user voice input into text, means for analyzing user order information using the interactive AI model and providing menu options, and means for tracking delivery status for a specific order and providing the user with an estimated delivery time. This makes it possible to streamline a series of operations, from order taking to delivery tracking and customer support, even in the food delivery industry.

[1627] "Specific industry" refers to all businesses related to a specific business area or field.

[1628] "Data collection" refers to the process of obtaining information relevant to a particular industry.

[1629] "Conversational artificial intelligence model" refers to a machine learning model that is trained to respond appropriately to user input in natural language.

[1630] "A company's business processes" refers to the set of procedures and methods by which a company carries out its daily business operations.

[1631] "Jargon" refers to words and phrases specific to a particular industry or field.

[1632] "Customization" refers to the process of tailoring a system or service to meet specific needs and requirements.

[1633] "Input Data" refers to information or instructions provided by a user.

[1634] "Dialogue log" refers to a record of all conversations exchanged between a user and an interactive AI system.

[1635] "Data analysis" refers to the process of analyzing accumulated data and extracting meaningful information and patterns.

[1636] "Model retraining" refers to the process of retraining an existing machine learning model based on new data or analytical results.

[1637] "Dialogue package" refers to a software module that forms part of a dialogue-based artificial intelligence system optimized for a specific industry.

[1638] "Voice input" refers to the process of capturing user spoken words into the system.

[1639] "Menu options" refer to selectable products and services offered to a user.

[1640] "Delivery status" refers to the progress of an ordered item until it is delivered.

[1641] "Expected Delivery Time" means the estimated time it will take for an ordered item to be delivered.

[1642] The present invention provides an interactive artificial intelligence system optimized for a specific industry, streamlining order acceptance, delivery tracking, and customer support in the food delivery industry. This system is configured to operate in cooperation with each element: server, terminal, and user.

[1643] server

[1644] The server provides the following means:

[1645] 1. Data collection method: Collect information related to a specific industry from external databases to understand industry-specific terminology and business processes.

[1646] 2. Model training method: Based on the collected data, a conversational artificial intelligence model (generative AI model) is trained, which builds a model that can respond appropriately to industry-specific questions and requests.

[1647] 3. Data analysis: Analyze the interaction logs sent by users to improve the accuracy of the model and extract new insights.

[1648] 4. Package development tools: Based on the analysis results, we will develop new industry-specific interactive packages to optimize order taking, delivery tracking, and customer support in the food delivery industry.

[1649] Terminal

[1650] The terminal provides the means to:

[1651] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system, allowing the system to be customized to meet the company's specific needs.

[1652] 2. Data reception and transmission means: Data input from the user (e.g., order details or questions) is transmitted to the server in real time, and response data from the server is presented to the user.

[1653] 3. Voice input means: Provides a voice recognition function to convert voice input into text.

[1654] 4. Package update means: Receive new industry-specific dialogue packages (generative AI models) distributed from the server and update the system.

[1655] User

[1656] The user provides the means to:

[1657] 1. Input method: Food delivery orders and inquiries are input into the conversational AI system. For example, you can say, "I'd like to order one Margherita pizza and one Coke Zero" by voice.

[1658] 2. Feedback means: Providing feedback on the system's response, which improves the accuracy of the system's response.

[1659] Hardware and software used

[1660] Hardware: Cloud servers (AWS, Google Cloud, etc.), smartphones (iOS, Android, etc.)

[1661] Software: Generative AI model (facebook / blenderbot-400M-distill), framework (Flask), speech recognition API (Google Cloud Speech-to-Text), database (MongoDB)

[1662] Data processing and calculation

[1663] Speech-to-text conversion: Uses the Google Cloud Speech-to-Text API to convert speech to text, allowing you to capture user speech as text.

[1664] Dialogue model response generation: A generative AI model trained on collected data is used to generate appropriate responses to user input. For example, if a user inputs, "I'd like to order one Margherita pizza and one Coke Zero," the conversational AI system will provide corresponding menu options.

[1665] Delivery status tracking and provision: Real-time delivery status information for specific orders is tracked and provided to users. For example, if a user requests "What is the delivery status of order number 12345?", the current delivery status and estimated delivery time will be provided.

[1666] Specific examples

[1667] Order Taker: "I'd like to order one Margherita pizza and one Coke Zero."

[1668] Delivery tracking: "What's the delivery status of order number 12345?"

[1669] Customer Support: "Why can't I use my coupon code?"

[1670] Prompt Sentence Examples

[1671] Enter your order: "I'd like to order one Margherita pizza and one Coke Zero."

[1672] Delivery tracking: "What's the delivery status of order number 12345?"

[1673] Support Enquiry: "Why can't I use my coupon code?"

[1674] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1675] Step 1:

[1676] Users voice-order food delivery through a smartphone application, where the voice input is captured.

[1677] Input: User speaks "I would like to order one Margherita pizza and one Coke Zero."

[1678] Output: Audio input data

[1679] Step 2:

[1680] The device converts voice input into text using the Google Cloud Speech-to-Text API.

[1681] Input: Voice input data

[1682] Output: Text data "I would like to order one Margherita pizza and one Coke Zero."

[1683] Specific operation: Call the voice recognition API and obtain the voice data as text data.

[1684] Step 3:

[1685] The device sends text data to the server, which uses a generative AI model (facebook / blenderbot-400M-distill) to generate an appropriate response to the user's order.

[1686] Input: Text data "I would like to order one Margherita pizza and one Coke Zero."

[1687] Output: Response data containing menu options

[1688] Specific operation: Text data is input into an interactive artificial intelligence model and response text is generated.

[1689] Step 4:

[1690] The server transmits the generated response data to the terminal, which then presents it to the user.

[1691] Input: Response data "Would you like one Margherita pizza and one Coke Zero?"

[1692] Output: The response text to display to the user

[1693] Specific behavior: Display the response text on the device's user interface.

[1694] Step 5:

[1695] The user provides confirmation, for example, a voice or text reply of "Yes, please."

[1696] Input: User response "Yes, please"

[1697] Output: Input data for verification

[1698] Step 6:

[1699] The terminal performs voice recognition again and sends the result as text data to the server. The server receives this confirmation data and processes it to confirm the order.

[1700] Input: Input data for confirmation "Yes, please"

[1701] Output: Order confirmation data

[1702] Specific operation: Converts speech into text and sends the data to the server to confirm the order. The server stores the order information in a database.

[1703] Step 7:

[1704] After the order is confirmed, the server generates a tracking ID for tracking the delivery status and sends it to the terminal. The terminal displays the tracking ID to the user.

[1705] Input: Order confirmation data

[1706] Output: Tracking ID

[1707] Specific operation: A tracking ID is generated based on the order information and sent to the terminal, which then displays the tracking ID on the user interface.

[1708] Step 8:

[1709] A user queries the application to check the delivery status of an order, entering something like "What is the delivery status of order number 12345?"

[1710] Input: User inquiry: "What is the delivery status of order number 12345?"

[1711] Output: Delivery tracking request data

[1712] Step 9:

[1713] The server tracks the delivery status, obtains the status in real time, and transmits the data to the terminal, which then presents the status data to the user.

[1714] Input: Delivery tracking request data

[1715] Output: Delivery status data "Current delivery status is on its way and will arrive in 15 minutes."

[1716] Specific operation: Obtain the current status from the delivery tracking system and send the information to the terminal.

[1717] Step 10:

[1718] The order is completed when the user receives it. The user provides feedback to the system, for example, by typing "The item arrived safely."

[1719] Input: User feedback "The item arrived safely."

[1720] Output: Feedback data

[1721] Specific operation: Feedback data is sent to the server and used to improve the system.

[1722] 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.

[1723] ---

[1724] This invention is a system that provides more sophisticated responses by combining a conversational AI system optimized for a specific industry with an emotion engine that recognizes the user's emotions. This system is configured so that the server, terminal, and user elements work together and the emotion engine is used to realize responses based on the user's emotional state.

[1725] server

[1726] The server provides the following means:

[1727] 1. Data collection methods: Data related to a specific industry are collected from external databases and specialized books.

[1728] 2. Model training method: Train an interactive artificial intelligence model based on the collected data.

[1729] 3. Emotion engine: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response.

[1730] 4. Data analysis: Analyze the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights.

[1731] 5. Package development method: Based on the analysis results, new industry-specific dialogue packages will be developed.

[1732] Terminal

[1733] The terminal provides the means to:

[1734] 1. Customization methods: Provide an interface for inputting a company's business processes and terminology into the system.

[1735] 2. Data reception and transmission means: Input data and emotion data from the user are transmitted to the server in real time, and response data from the server is presented to the user.

[1736] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system.

[1737] User

[1738] The user provides the means to:

[1739] 1. Input means: Use conversational artificial intelligence in your daily work by inputting questions and commands into the system.

[1740] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This data is sent to the emotion engine.

[1741] 3. Feedback measures: Providing feedback on the system's response.

[1742] Example: Emotion recognition AI system for medical institutions

[1743] As a specific example, an embodiment of an emotion-recognition conversational artificial intelligence system for medical institutions will be described below:

[1744] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[1745] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[1746] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[1747] Users (doctors and nurses) use conversational AI in their daily medical practice. For example, when a doctor inputs "Please tell me about patient X's allergies," the device sends this input data and the patient's emotional data to the server.

[1748] The server generates an appropriate and emotionally sensitive response based on the user input and emotion data and sends it to the terminal.

[1749] The terminal presents the generated response to the user.

[1750] The server collects and analyzes user-AI dialogue logs and emotional data, and uses the results of this analysis to retrain the model and develop a new dialogue package specialized for medical work.

[1751] In this way, by combining a conversational AI system optimized for a specific industry with an emotion engine, the present invention can provide more precise and attentive responses to users, significantly improving business efficiency for companies.

[1752] The processing flow will be explained below.

[1753] ---

[1754] Step 1:

[1755] The server collects data related to a particular industry from external databases and specialized books.

[1756] Step 2:

[1757] The server cleanses the collected data and converts it into a format that can be applied to AI models.

[1758] Step 3:

[1759] The server uses the cleansed data to train a basic interactive artificial intelligence model.

[1760] Step 4:

[1761] The server incorporates an emotion engine that recognizes the user's emotions, allowing it to analyze the emotion data and reflect it in responses.

[1762] Step 5:

[1763] The terminal is installed in a company's system and provides an interface for inputting company-specific business processes and terminology.

[1764] Step 6:

[1765] Users input their company's specific business processes and terminology through the interface.

[1766] Step 7:

[1767] The server collects input data from users and reflects it in the AI ​​model.

[1768] Step 8:

[1769] Users use conversational AI in their daily work, utilizing cameras and microphones to capture emotional data, for example by typing, "What's on the agenda for today's meeting?"

[1770] Step 9:

[1771] The terminal transmits the user's input data and emotion data to the server in real time.

[1772] Step 10:

[1773] The server generates a response based on the user's input and emotion data and sends the response to the terminal.

[1774] Step 11:

[1775] The terminal presents the generated response to the user.

[1776] Step 12:

[1777] The server securely stores the dialogue logs and emotional data between the user and the AI.

[1778] Step 13:

[1779] The server periodically retrieves the accumulated dialogue logs and emotion data and performs data preprocessing.

[1780] Step 14:

[1781] The server analyzes the preprocessed data and sentiment data to identify areas for improvement in the model.

[1782] Step 15:

[1783] The server retrains the conversational AI model based on the analysis results.

[1784] Step 16:

[1785] The server develops new industry-specific dialogue packages.

[1786] Step 17:

[1787] The terminal receives new interaction packages from the server and updates the system.

[1788] Step 18:

[1789] A user uses the new interaction package to perform a daily task, for example, "What is customer A's past purchase history?"

[1790] Step 19:

[1791] The device sends the user's input data and emotional data to the server, and the conversational AI generates an appropriate response and returns it to the device.

[1792] Step 20:

[1793] The terminal presents the new response to the user.

[1794] Step 21:

[1795] The server continues to store updated dialogue logs and emotional data to help improve the accuracy of the system.

[1796] ---

[1797] These are the specific steps in programming a conversational AI system that combines an emotion engine and is optimized for a specific industry.

[1798] Example 2

[1799] 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."

[1800] Conventional conversational AI systems generate responses only in response to user input and do not take into account the user's emotional state, which can result in a lack of appropriate responses and consideration. Furthermore, generating responses based on the user's emotional state is difficult, which reduces the accuracy and reliability of systems specialized for specific industries.

[1801] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for recognizing and acquiring user emotion data, means for receiving input data and emotion data from the user and analyzing them in real time, means for generating an appropriate response based on the analyzed data, means for presenting the generated response to the user, means for analyzing accumulated dialogue logs and emotion data and retraining the model, means for developing and updating a new industry-specific dialogue package, and means for providing the updated dialogue package to the user. This makes it possible to generate responses according to the user's emotional state, thereby significantly improving business efficiency and reliability in a specific industry.

[1802] "Means for collecting data related to a specific industry" refers to systems or software for automatically collecting information and data related to a specific industry from external databases or specialized books.

[1803] A "means for training an interactive artificial intelligence model" is a method for training an interactive artificial intelligence model using a machine learning algorithm based on collected data.

[1804] "Means for recognizing and acquiring user emotional data" refers to technology that recognizes a user's facial expressions and tone of voice through devices such as a camera or microphone, and acquires their emotional state as data.

[1805] "Means for receiving input data and emotional data from a user and analyzing it in real time" refers to a system in which text, voice data, and emotional data sent by a user are instantly received by a server and analyzed.

[1806] The "means for generating an appropriate response based on the analyzed data" is a method for generating an appropriate response based on user input and emotion data analyzed in real time.

[1807] The "means for presenting the generated response to the user" is a system that displays or audibly conveys the response generated by the server to the user via the terminal.

[1808] "Means for analyzing accumulated dialogue logs and emotional data and retraining a model" refers to a method for analyzing dialogue history and emotional data, and retraining an interactive artificial intelligence model based on the analysis to improve its performance.

[1809] "Means for developing and updating new industry-specific dialogue packages" refers to technology that creates dialogue scenarios and response models optimized for specific industries based on analysis results and new data, and updates the system.

[1810] The "means for providing an updated dialogue package to a user" refers to a method for delivering a newly developed or updated dialogue artificial intelligence package to a user's terminal and making it available for use.

[1811] This invention is a system that provides more accurate responses by combining an emotional engine with a conversational AI system optimized for a specific industry. This system works in conjunction with the server, terminal, and user elements to realize responses based on the user's emotional state.

[1812] Server Features

[1813] The server provides the following main functions:

[1814] 1. Data collection method: The server collects data related to a specific industry from external databases and specialized books. For example, the server uses the PubMed API to obtain medical-related data.

[1815] 2. Model training method: Train an interactive AI model based on the collected data. This method uses machine learning frameworks such as TensorFlow and PyTorch. For example, build an interactive model of the medical treatment process based on medical data.

[1816] 3. Emotion Engine: The server uses emotion recognition software such as Emotion API to analyze the user's facial expressions and tone of voice and obtain emotional data, which is then reflected in the generation of a response.

[1817] 4. Data analysis: Analyzes the dialogue logs and sentiment data sent by users to improve the accuracy of the model and extract new insights, thereby improving the efficiency and accuracy of the entire system.

[1818] 5. Package development: Based on the analysis results, new industry-specific dialogue packages are developed and the system is updated. This method is important for generating new dialogue scenarios and response models.

[1819] Device Features

[1820] The terminal provides the following main features:

[1821] 1. Customization: Provides an interface for inputting a company's business processes and terminology into the system. For example, in a hospital system, it sets the terminology commonly used by doctors and nurses.

[1822] 2. Data reception and transmission means: Sends input data and emotion data from the user to the server in real time, and presents the response from the server to the user. For this function, for example, the Zoom SDK can be used.

[1823] 3. Package update means: Receives new industry-specific dialogue packages distributed from the server and updates the system, ensuring that the latest data and response models are always available.

[1824] User operations

[1825] The user does the following:

[1826] 1. Input Method: Conversational AI is used in daily operations to input questions or commands into the system. For example, a doctor might type, "Please tell me Patient X's allergy information." This input can be done via keyboard, voice, or touch panel.

[1827] 2. Emotion recognition means: Emotion data is acquired from the user's facial expressions and tone of voice through devices such as cameras and microphones. This acquired emotion data is sent to the emotion engine on the server.

[1828] 3. Feedback measures: Providing feedback on the system's response. This feedback helps refine and improve the system.

[1829] Specific examples

[1830] For example, an emotion-aware conversational AI system for a medical institution:

[1831] The server collects data on medical terminology and medical treatment processes from medical databases and trains an interactive artificial intelligence model.

[1832] The emotion engine recognizes the emotional state of patients and doctors, analyzes that data, and reflects it in responses.

[1833] The device will connect to the hospital's systems, provide an interface for doctors and nurses to set commonly used terms and medical procedures, and will also be equipped with a device that captures the patient's emotional state through a camera and microphone.

[1834] Users (doctors and nurses) use conversational artificial intelligence in their daily medical practice.

[1835] Prompt Sentence Examples

[1836] "Please tell me Patient X's allergy information and current emotional state."

[1837] "Please advise on a treatment process that takes into consideration the patient's feelings."

[1838] "Please provide examples of responses for specific medical conditions."

[1839] This invention combines an interactive artificial intelligence system optimized for a specific industry with an emotion engine, which is expected to realize highly accurate responses that are attuned to the user, significantly improving the efficiency and reliability of business operations.

[1840] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1841] Step 1: Data collection

[1842] The server collects data related to a specific industry from external databases and specialized books. The input for this process is a specific keyword or query, and the output is related data. For example, in a system for medical institutions, the server uses the PubMed API to automatically collect literature data based on keywords such as "allergy" and "medical treatment process." The collected data is then used in subsequent model training.

[1843] Step 2: Model training

[1844] The server trains an interactive AI model based on the collected data. The input for this process is the collected data, and the output is a trained interactive AI model. Specifically, it uses machine learning frameworks such as TensorFlow and PyTorch to build and train dialogue scenarios based on medical treatment processes using medical data. The trained model is then used to generate responses to questions from users.

[1845] Step 3: User Input

[1846] The user inputs a question or command into the system. The input for this process is text or voice input from the user, and the output is the input data. For example, a doctor might input "Please tell me about patient X's allergy information" into a terminal. This input data is used in the subsequent emotion recognition and response generation steps.

[1847] Step 4: Emotion Recognition

[1848] The device acquires emotion data from the user's facial expressions and tone of voice via a camera and microphone. The input for this process is the user's visual and audio characteristics, and the output is emotion data. Specifically, OpenCV is used to analyze the doctor's facial expressions captured by the camera to determine whether the user is feeling stressed. The emotion data is sent to the server and used in the subsequent response generation step.

[1849] Step 5: Response Generation

[1850] The server generates an appropriate response based on the user's input data and emotional data. The input for this process is the user's question or command, as well as emotional data, and the output is a response based on that. Specifically, using a generative AI model (e.g., GPT-4), if a user inputs "Tell me about patient X's allergies," the server generates a response based on the emotional data, such as "Patient X's allergy information is ____. Also, a high stress level has been detected, so special care is required."

[1851] Step 6: Present the response

[1852] The terminal presents the generated response to the user. The input for this process is the response data received from the server, and the output is the content presented to the user. Specifically, the terminal's display and voice output function are used to notify the doctor, "Patient X's allergy information is ____. Also, it appears that his current stress level is high, so he requires special attention."

[1853] Step 7: Data analysis and model updating

[1854] The server collects and analyzes user-AI dialogue logs and emotional data. The inputs for this process are dialogue logs and emotional data, and the output is the analysis results and an improved dialogue model. Specifically, the accuracy of the model is improved and new insights are extracted based on the dialogue logs and emotional data, and the model is retrained using TensorFlow. The analysis results are used to develop new industry-specific dialogue packages and update the system.

[1855] (Application example 2)

[1856] 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."

[1857] Modern brick-and-mortar stores require effective methods to improve customer shopping experiences. However, conventional methods struggle to grasp customers' emotional states in real time and provide appropriate product recommendations. Furthermore, conversational AI systems generate responses without considering customers' emotions, resulting in low customer satisfaction. Therefore, it is important to provide personalized product recommendations based on customers' emotional states.

[1858] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data related to a specific industry, means for training an interactive artificial intelligence model using the collected data, means for customizing the company's business processes and terminology, emotion engine means for recognizing the user's emotional state, analyzing the emotion data, and reflecting the emotion data in responses, and data processing means for providing product suggestions based on the customer's emotional state. This makes it possible to provide personalized product suggestions based on the customer's emotional state in real time.

[1859] "Specific industry" refers to a specific commercial or business sector that specializes in handling data and business processes related to that industry.

[1860] "Data collection means" refers to the means of obtaining information and data related to a specific industry from external databases and materials.

[1861] An "interactive artificial intelligence model" refers to an artificial intelligence algorithm or program that enables natural dialogue with the user.

[1862] A "model training means" is a means for training an interactive artificial intelligence model using collected data to improve its performance.

[1863] "Business process" refers to a series of procedures and flows that a company follows when carrying out its business.

[1864] "Jargon" refers to specialized words and phrases used in a particular industry or field.

[1865] "Customization tools" refers to configuration tools and interfaces that allow a company to adapt the system to its specific business processes and terminology.

[1866] "Input data" refers to information and instructions provided by a user to a system.

[1867] "Interaction log" refers to a record of interaction data between a user and a system.

[1868] "Analytical tools" refers to techniques and methods for analyzing accumulated data and using the results to improve the model.

[1869] "Emotion engine means" refers to technology or algorithms for recognizing the user's emotional state, analyzing that data, and reflecting it in dialogue responses.

[1870] "Data processing means" refers to a mechanism for analyzing acquired data and performing specific processing.

[1871] "Camera and microphone means" refers to a photographic and audio recording device for capturing the customer's facial expressions and tone of voice in real time.

[1872] "Real-time" means near-simultaneous data processing and response.

[1873] "Personalized product proposals" refer to proposals for products and services that are individually customized based on the customer's individual attributes and circumstances.

[1874] The present invention is a system that provides personalized product recommendations using a conversational AI system combined with an emotion engine that recognizes user emotions. This system collects data related to a specific industry, trains a conversational AI model, and customizes the company's business processes and terminology to provide optimized services to customers.

[1875] server

[1876] The server provides the following means:

[1877] 1. Data collection methods: Collect data related to a specific industry from external databases and sources. For example, in the distribution industry, collect and use consumer behavior data.

[1878] 2. Model training method: The collected data is used to train a conversational artificial intelligence model. The model used here is a generative AI model that learns from a large amount of conversational data.

[1879] 3. Emotion engine means: Recognizes the user's emotions, analyzes the emotional data, and reflects it in the response. This engine uses the DeepFace library to recognize emotions from facial expressions and tone of voice.

[1880] 4. Data processing means: Processes data to provide product suggestions based on the customer's emotional state. This means obtains product recommendations using external APIs.

[1881] 5. Analysis method: Preprocess the accumulated dialogue logs and perform data analysis, which allows us to retrain the model and improve its accuracy.

[1882] Terminal

[1883] The terminal provides the means to:

[1884] 1. Customization: Provides an interface for setting up a company's business processes and terminology. For example, inputting terminology specific to the distribution industry.

[1885] 2. Input data receiving means: Provides a communication means for transmitting input data from the user to the server in real time.

[1886] 3. Emotion recognition means: Equipped with a camera and microphone to capture the customer's facial expressions and tone of voice in real time. For example, the camera and microphone of a smartphone can be used to recognize the customer's emotions.

[1887] 4. Response providing means: Provides the product suggestions sent from the server to the user. For example, displays the acquired product data on the smartphone screen.

[1888] User

[1889] The user provides the means to:

[1890] 1. Input method: Customers input their questions or opinions using conversational AI. For example, they can input "What's your recommendation today?" through a smartphone app.

[1891] 2. Emotion recognition: Emotion data is acquired from the user's facial expressions and tone of voice via a camera and microphone. This data is sent to the server in real time.

[1892] 3. Feedback measures: Providing feedback on the system's response, e.g., rating satisfaction with a suggested product.

[1893] Specific examples

[1894] For example, if the user is feeling stressed, the system will suggest relaxation-related products, and if the user is smiling, entertainment products that will increase their enjoyment.

[1895] Prompt Sentence Examples

[1896] "The emotion the customer is feeling is stress. Make product recommendations based on this emotion."

[1897] In this way, by using an interactive artificial intelligence system combined with an emotion engine, the present invention can provide personalized services based on the customer's emotional state in real time, thereby improving the shopping experience in physical stores.

[1898] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1899] Step 1:

[1900] The device captures the customer's facial expressions and tone of voice using a camera and microphone. The input is the customer's facial expression data and voice data. This data is collected in real time and saved as image and audio files.

[1901] Step 2:

[1902] The device sends the collected facial expression data and voice data to the server. The input is the image and voice files collected in step 1. Uploading this to the server enables emotion analysis.

[1903] Step 3:

[1904] The server analyzes the facial expression data using the DeepFace library to identify the customer's emotional state. The input is the image file sent in step 2. This data is passed to DeepFace for processing and the output is the emotion analysis result.

[1905] Step 4:

[1906] The server generates a prompt to suggest products appropriate for the customer based on their emotional state and sends a request to the external API. The input is the emotion analysis result obtained in step 3. Based on this, the server generates a prompt such as "The customer is feeling ____. Please suggest products based on this emotion," and sends it to the external API.

[1907] Step 5:

[1908] An external API generates product suggestions based on the prompt sent. The input is the prompt sent in step 4. The API parses it and generates a list of relevant products. The output is the data of the recommended products.

[1909] Step 6:

[1910] The server receives the product suggestions obtained from the external API. The input is the product list data generated in step 5. This data is stored in the server and prepared for transmission to the device.

[1911] Step 7:

[1912] The terminal displays the product proposals received from the server to the customer. The input is the product list data received in step 6. This data is displayed on the smartphone screen, providing specific product information to be proposed to the customer.

[1913] Step 8:

[1914] The user provides feedback on the proposed product. The input is the product information displayed in step 7. The user enters their satisfaction level and opinions about the product and sends them to the server via their terminal.

[1915] Step 9:

[1916] The server analyzes the collected feedback data and uses it to retrain the AI ​​model. The input is the feedback data obtained in step 8. The output is a new and improved AI model.

[1917] In this way, a system is completed that provides personalized product suggestions based on the user's emotional state.

[1918] 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.

[1919] 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.

[1920] 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.

[1921] 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.

[1922] 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.

[1923] 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.

[1924] 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).

[1925] 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.

[1926] 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."

[1927] 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.

[1928] 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).

[1929] 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.

[1930] 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.

[1931] 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.

[1932] 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.

[1933] 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.

[1934] 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.

[1935] 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.

[1936] 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.

[1937] 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.

[1938] 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.

[1939] The following is further disclosed regarding the above embodiment.

[1940] ---

[1941] (Claim 1)

[1942] A means of collecting data related to a particular industry;

[1943] means for training an interactive artificial intelligence model using the collected data;

[1944] A means to customize your company's business processes and terminology,

[1945] means for receiving input data from a user and storing an interaction log;

[1946] A means of analyzing the accumulated dialogue logs and retraining the model;

[1947] A means to develop and update new industry-specific dialogue packages;

[1948] means for providing an updated interaction package to a user;

[1949] A system including:

[1950] (Claim 2)

[1951] The system according to claim 1, further comprising a means for inputting business processes and technical terms of a company.

[1952] (Claim 3)

[1953] 2. The system according to claim 1, further comprising means for preprocessing the accumulated dialogue logs and performing data analysis.

[1954] "Example 1"

[1955] (Claim 1)

[1956] A means of collecting data related to a particular industry;

[1957] means for training an interactive artificial intelligence model using the collected data;

[1958] A means to customize your company's business processes and terminology,

[1959] means for receiving input data from a user and transmitting it to a server in real time;

[1960] means for generating a response to a user;

[1961] A means of analyzing the accumulated dialogue logs and retraining the model;

[1962] A means to develop and update new industry-specific dialogue packages based on the analyzed feedback; and

[1963] means for delivering updated dialogue packages to the terminals and automatically updating the system;

[1964] A system including:

[1965] (Claim 2)

[1966] 2. The system according to claim 1, further comprising means for providing an interface for inputting business processes and technical terms of a company.

[1967] (Claim 3)

[1968] 2. The system according to claim 1, further comprising means for preprocessing the accumulated dialogue logs and performing data analysis.

[1969] "Application Example 1"

[1970] (Claim 1)

[1971] A means of collecting data related to a particular industry;

[1972] means for training an interactive artificial intelligence model using the collected data;

[1973] A means to customize your company's business processes and terminology,

[1974] means for receiving input data from a user and storing an interaction log;

[1975] A means of analyzing the accumulated dialogue logs and retraining the model;

[1976] A means to develop and update new industry-specific dialogue packages;

[1977] means for providing an updated interaction package to a user;

[1978] means for converting voice input from a user into text;

[1979] means for analyzing a user's order information using an interactive artificial intelligence model and providing menu options;

[1980] a means for tracking delivery status for a particular order and providing the user with an estimated delivery time;

[1981] A system including:

[1982] (Claim 2)

[1983] The system according to claim 1, further comprising a means for inputting business processes and technical terms of a company.

[1984] (Claim 3)

[1985] 2. The system according to claim 1, further comprising means for preprocessing the accumulated dialogue logs and performing data analysis.

[1986] "Example 2: Combining Emotion Engines"

[1987] (Claim 1)

[1988] A means of collecting data related to a particular industry;

[1989] means for training an interactive artificial intelligence model using the collected data;

[1990] means for recognizing and acquiring user emotion data;

[1991] means for receiving input data and emotion data from a user and analyzing them in real time;

[1992] means for generating an appropriate response based on the analyzed data;

[1993] means for presenting the generated response to a user;

[1994] a means for analyzing the accumulated dialogue logs and sentiment data and retraining the model;

[1995] A means to develop and update new industry-specific dialogue packages;

[1996] means for providing an updated interaction package to a user;

[1997] A system including:

[1998] (Claim 2)

[1999] 2. The system according to claim 1, further comprising means for customizing a company's business processes and terminology.

[2000] (Claim 3)

[2001] 2. The system according to claim 1, further comprising means for preprocessing the accumulated dialogue log and emotion data and performing data analysis.

[2002] "Application example 2 when combining emotion engines"

[2003] Claiming a new invention

[2004] (Claim 1)

[2005] A means of collecting data related to a particular industry;

[2006] means for training an interactive artificial intelligence model using the collected data;

[2007] A means to customize your company's business processes and terminology,

[2008] means for receiving input data from a user and storing an interaction log;

[2009] A means of analyzing the accumulated dialogue logs and retraining the model;

[2010] A means to develop and update new industry-specific dialogue packages;

[2011] an emotion engine means for recognizing an emotional state of a user, analyzing the emotional data, and reflecting the data in a response;

[2012] data processing means for providing product suggestions based on the emotional state of the customer;

[2013] a camera and microphone means for capturing the customer's facial expressions and tone of voice in real time;

[2014] A system including:

[2015] (Claim 2)

[2016] 2. The system according to claim 1, further comprising a means for inputting business processes and technical terms of a company.

[2017] (Claim 3)

[2018] 2. The system according to claim 1, further comprising means for preprocessing the accumulated dialogue logs and performing data analysis.

[2019] keyword

[2020] Generative AI model, prompt sentence [Explanation of symbols]

[2021] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting data related to a particular industry; means for training an interactive artificial intelligence model using the collected data; A means to customize your company's business processes and terminology, means for receiving input data from a user and storing an interaction log; A means of analyzing the accumulated dialogue logs and retraining the model; A means to develop and update new industry-specific dialogue packages; means for providing an updated interaction package to a user; A system including:

2. 2. The system according to claim 1, further comprising means for inputting business processes and technical terms of a company.

3. 2. The system according to claim 1, further comprising means for preprocessing the accumulated dialogue logs and performing data analysis.

Images