System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting historical behavior, books, video recordings, etc. of a business operator; means for cleaning and formatting the collected information; means for training a generative AI model using the pre-processed information; means for generating business decisions based on a specific business situation or scenario; means for providing recommendations using the generative AI model for user business consultation; and means for updating the generative AI model using new information or feedback.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Current corporate management faces a lack of methods for passing on the knowledge and experience of outstanding managers. In particular, when a prominent manager suddenly retires or passes away, it becomes difficult to share that knowledge and experience, significantly impacting the company's management decisions. Another notable problem is the lack of an efficient platform through which external managers and entrepreneurs can receive prompt and appropriate management advice. Given this background, there is a need for a method that can provide management advice widely while properly passing on the knowledge of managers. [Means for solving the problem]

[0005] In order to solve the above problems, the present invention provides a system that includes the following means. First, a means is provided for collecting a manager's past words and actions, books, video recordings, etc. Next, a means is provided for cleaning the collected data and standardizing the format. This provides a means for training a generative AI model using preprocessed data. Furthermore, a means is provided for generating business decisions based on specific business situations and scenarios, and for providing advice using the generative AI model in response to a user's business consultation. A means is also provided for updating the generative AI model using new data and feedback. These means enable the inheritance of the manager's knowledge and the provision of efficient and appropriate business decisions and consultations.

[0006] A "manager" is a person who leads a company or organization and is in a position to decide management policies and strategies.

[0007] "Words and actions" refers to the general statements and actions made by an individual, and reflects that person's thoughts and standards of judgment.

[0008] A "book" is a collection of documents, in print or electronic format, that contains information or knowledge on a particular subject.

[0009] "Visual recording" refers to visual data stored in the form of video or film using a camera or other photographic device.

[0010] "Data collection means" refers to a method or device for obtaining and accumulating necessary information from the Internet or a database.

[0011] "Cleaning" refers to the process of removing noise and unnecessary information from data and arranging it into a useful format.

[0012] "Unifying formats" refers to the process of converting different types and formats of data into a single standard format to ensure consistency.

[0013] A "generative AI model" is a type of artificial intelligence that refers to an algorithm trained to generate specific patterns or decisions from input data.

[0014] "Training" refers to the process of using data to enhance an AI model and improve its performance.

[0015] "Management decisions" refer to important decisions made in the operation of a company or organization, including strategy formulation, risk assessment, and resource allocation.

[0016] "Management consulting" refers to the act of providing advice and solutions to specific management problems and issues.

[0017] "Feedback" refers to the reaction or evaluation received to the output of a system or process, which is used to make subsequent improvements or adjustments.

[0018] "Update" is the act of bringing a system or model up to date with new information or conditions.

[0019] "Database" refers to a system or location for storing and managing systematically collected information or data.

[0020] "User terminal" refers to a device that a user uses to access and operate the system, including a PC or smartphone. [Brief explanation of the drawings]

[0021] [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

[0022] 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.

[0023] First, the terms used in the following description will be explained.

[0024] 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).

[0025] 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.

[0026] 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.

[0027] 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.

[0028] 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."

[0029] [First embodiment]

[0030] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0031] 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.

[0032] 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).

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0038] 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.

[0039] 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.

[0040] 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.

[0041] 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."

[0042] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[0043] Data collection

[0044] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[0045] Data Preprocessing

[0046] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[0047] Model training

[0048] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[0049] Business decision simulation

[0050] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[0051] Management Consulting Platform

[0052] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[0053] Model Update

[0054] The server periodically updates the generative AI model using new data and feedback, for example by collecting newly acquired executive interviews or speeches and adding them to the existing dataset, ensuring that the generative AI model always provides decisions based on the most up-to-date information.

[0055] Specific examples

[0056] Examples of data collection

[0057] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0058] Specific examples of business decision simulation

[0059] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0060] Examples of management consulting platforms

[0061] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0062] The above is an embodiment of the present invention. By using this system, it becomes possible to inherit the knowledge of the manager and to make efficient and appropriate management decisions and provide management consultations.

[0063] The processing flow will be explained below.

[0064] Data collection

[0065] Step 1:

[0066] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[0067] Step 2:

[0068] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[0069] Step 3:

[0070] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[0071] Data Preprocessing

[0072] Step 1:

[0073] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[0074] Step 2:

[0075] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[0076] Step 3:

[0077] The server converts the cleaned data into a unified format and stores it in a database.

[0078] Model training

[0079] Step 1:

[0080] The server splits the preprocessed dataset into a training set and a test set.

[0081] Step 2:

[0082] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[0083] Step 3:

[0084] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[0085] Step 4:

[0086] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[0087] Business decision simulation

[0088] Step 1:

[0089] The terminal provides an interface to the simulation tool for internal users.

[0090] Step 2:

[0091] The user inputs a specific business scenario (e.g., new market entry, product development).

[0092] Step 3:

[0093] The server receives the input scenario and queries the generative AI model.

[0094] Step 4:

[0095] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[0096] Step 5:

[0097] The server returns the results to the terminal, where the user views the results.

[0098] Management Consulting Platform

[0099] Step 1:

[0100] The terminal provides an interface for the management consulting platform to external users.

[0101] Step 2:

[0102] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[0103] Step 3:

[0104] The server receives the input consultation content and executes a query against the generative AI model.

[0105] Step 4:

[0106] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[0107] Step 5:

[0108] The server sends advice and solutions to the terminal, which the user then confirms.

[0109] Model Update

[0110] Step 1:

[0111] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[0112] Step 2:

[0113] The server cleans the new data and standardizes the format.

[0114] Step 3:

[0115] The server retrains the generative AI model with the updated data.

[0116] Step 4:

[0117] The server evaluates the performance of the retrained model and releases an update.

[0118] The above are the specific steps of the system program processing.

[0119] Example 1

[0120] 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."

[0121] There is a need for a system that can automatically and efficiently provide appropriate advice by systematizing management decisions that reflect the past words, actions, and thoughts of managers. However, current systems require cumbersome data collection and preprocessing, making it difficult to train effective generative AI models. Furthermore, users lack a means to instantly obtain appropriate management decisions and advice for specific scenarios. Furthermore, the process of updating the model based on new data and feedback is not automated, making it difficult to keep up with the latest management information.

[0122] 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.

[0123] In this invention, the server includes means for collecting the manager's past words and actions, literature, video recordings, etc., means for cleaning the collected data and standardizing the format, and means for training the generative AI model using the preprocessed data. This enables means for a user to input a specific management scenario and generate management decisions based on that scenario, means for providing advice using the generative AI model in response to a user's management consultation, and means for updating the generative AI model using new data and feedback.

[0124] A "management officer" is a person who is responsible for making important decisions in the operation of a company or organization.

[0125] "Behavior" refers collectively to speech and actions, and refers to words and actions shown in specific situations or on specific themes.

[0126] "Literature" is any kind of written material written to provide information, such as books, articles, or reports.

[0127] A "visual recording" is a collection of visual and audio information stored in the form of video, video recording or other media.

[0128] "Data collection" is the process of obtaining and gathering information needed for a specific purpose.

[0129] "Cleaning" is the process of removing errors and noise from collected data, organizing and normalizing the data.

[0130] "Unifying formats" is the process of converting data with different formats and structures into a consistent, standard format.

[0131] "Preprocessing" refers to data preparation performed before data analysis or model training.

[0132] A "generative AI model" is an algorithm that uses artificial intelligence techniques to learn specific patterns and characteristics and make inferences based on new data.

[0133] "Training" is the process of teaching a generative AI model using data to improve its performance.

[0134] A "business scenario" is a setting that represents assumptions and conditions regarding a particular business situation or strategy.

[0135] "Management decision-making" is the process by which managers make rational decisions regarding the operation of a company or organization.

[0136] "Management consultation" is the act of seeking professional advice on management issues and problems.

[0137] "Advice" is helpful advice or direction given on a specific problem or issue.

[0138] "Feedback" is evaluation or information provided to improve the performance of a system or model.

[0139] "Updating" is the process of adding new information or data to an existing system or model to improve its performance or accuracy.

[0140] A "database" is an organized collection of data that can be efficiently managed and searched.

[0141] A "user terminal" is a device or interface that a user can directly access and operate.

[0142] overview

[0143] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[0144] Data collection

[0145] The server first collects information on the manager's past statements and actions, documents, video recordings, etc. Tools used include web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). Furthermore, the video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and the document data is digitized using OCR technology (e.g., Tesseract).

[0146] Data Preprocessing

[0147] The server cleans the collected data and standardizes its format. The cleaning process involves removing duplicate data and removing noise. Tools used include NLP libraries (e.g., NLTK, spaCy). This data processing involves tokenizing the text data (splitting it into words) and storing it in a consistent format in the database.

[0148] Model training

[0149] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework such as TensorFlow or PyTorch. The dataset is divided into a training set and a test set, and the model's performance is evaluated and improved through cross-validation.

[0150] Business decision simulation

[0151] The terminal provides an interface that users can access and input specific business scenarios. For example, a web application may be provided that provides a form where users can input scenarios such as "entering a new market."

[0152] Users input scenarios through a simulation tool, and the server inputs prompt statements (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate business decisions.

[0153] The server sends the generated business decisions back to the terminal, allowing the user to view the results, for example by displaying the results on a dashboard so the user can view the details.

[0154] Management Consulting Platform

[0155] The terminal provides an interface for external users to receive management consultations. For example, it provides a chatbot and a consultation form, allowing users to input their consultation details, such as "fundraising strategies" or "recruitment methods."

[0156] The user inputs a specific business consultation, and the server generates advice using a generative AI model.

[0157] The server sends the generated advice back to the terminal, and the user makes management decisions based on that information. The advice is displayed in a chat window or in a report format.

[0158] Model Update

[0159] The server periodically updates the generative AI model with new data and feedback, for example by adding newly acquired executive interviews or speech data and updating the existing dataset.

[0160] The server retrains the model with the latest data to optimize its performance.

[0161] Specific examples

[0162] Examples of data collection

[0163] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0164] Specific examples of business decision simulation

[0165] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0166] Examples of management consulting platforms

[0167] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0168] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management.

[0169] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0170] Step 1: Data collection

[0171] The server collects data related to the manager. Input includes the manager's lecture videos, blog posts, interview articles, etc. from online platforms. Specific operations include using web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). These tools are used to automatically collect the necessary data and save it as a file. The output is raw data stored in local storage.

[0172] Step 2: Data Preprocessing

[0173] The server preprocesses the collected data. The input includes the raw data collected in step 1. Specifically, it converts voice data into text using speech recognition technology (e.g., Google Speech-to-Text API), and converts image and PDF data into digital text using OCR technology (e.g., Tesseract). It also performs noise removal, de-duplicate data removal, and tokenization using NLP libraries (e.g., NLTK, spaCy). The output is text data in a clean, unified format.

[0174] Step 3: Data storage

[0175] The server stores the preprocessed data in a database. The input includes the clean text data generated in step 2. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to store the text data in a structured format in the database. The output is the preprocessed text data stored in the database.

[0176] Step 4: Model training

[0177] The server uses the preprocessed data to train a generative AI model. The input includes the text data stored in the database in step 3. Specifically, a model is built using a deep learning framework (e.g., TensorFlow, PyTorch) and supervised learning and reinforcement learning are performed. The dataset is divided into a training set and a test set, and the model's performance is evaluated and optimized through cross-validation. The output is a generative AI model that can reproduce the words, actions, and thoughts of the manager.

[0178] Step 5: Business decision simulation

[0179] The terminal provides an interface for the user to input a business scenario. The input includes the business scenario (e.g., "Enter a new market") entered by the user. Specifically, it provides a web form or dashboard to allow the user to input the scenario. The server receives this scenario and inputs a prompt statement (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate a business decision. The output is the generated business decision, which is sent back to the terminal and displayed to the user.

[0180] Step 6: Management Consulting Platform

[0181] The terminal provides an interface for external users to input business advice. The input includes the specific advice entered by the user (e.g., "fundraising strategy"). Specific operations include providing a chatbot and a consultation form, allowing the user to input the details of the consultation. The server receives the consultation details and generates appropriate advice using a generative AI model. The output is the generated advice, which is sent back to the terminal and displayed to the user.

[0182] Step 7: Model Update

[0183] The server periodically updates the generative AI model using new data and feedback. The input includes newly collected managerial utterance data and feedback data. Specifically, it adds new data to the old dataset and retrains the model to optimize its performance. The output is an updated generative AI model, which can provide management decisions and advice based on the latest information.

[0184] (Application example 1)

[0185] 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."

[0186] In modern virtual store operations, managers must make a wide range of management decisions quickly and appropriately. However, relying on the manager's personal knowledge and experience can lead to biased decisions, resulting in insufficient risk assessment and strategic proposals. Furthermore, traditional management support systems have difficulty providing advice in real time, making them less effective in situations where immediate decisions are required on-site. This increases the burden on virtual store operators and hinders efficient operations, creating the challenge.

[0187] 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.

[0188] In this invention, the server includes means for collecting the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using the preprocessed data, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, means for providing management decision advice to the virtual store manager in real time via smart glasses, and means for generating risk assessments and strategic proposals based on scenarios specified by the user. This enables the virtual store manager to receive advice from the generative AI model in real time and make quick and appropriate management decisions.

[0189] A "management officer" is a person responsible for making decisions regarding the operation of a company or organization.

[0190] "Behavior" refers to all behavior, including verbal expressions and actions.

[0191] A "book" is a collection of bound documents that contain printed text and / or images and have a definite form.

[0192] "Visual recording" refers to data containing visual information in the form of video, film, etc.

[0193] "Collection methods" refer to methods and techniques for systematically gathering specific information or data.

[0194] "Cleaning" refers to the process of removing noise from data and arranging it into a unified format.

[0195] A "generative AI model" is an artificial intelligence model that learns using machine learning algorithms based on collected data.

[0196] "Training methods" refer to methods or techniques for training a generative AI model using a specific dataset.

[0197] "Business condition" refers to the current performance and environmental state of a company or organization.

[0198] A "scenario" refers to a hypothetical series of events or happenings based on specific circumstances or conditions.

[0199] "Advice delivery means" refers to the methods and technologies for communicating business decisions and proposals obtained by generative AI models to users.

[0200] "Update methods" refer to methods and techniques that incorporate new data and feedback into generative AI models to improve their accuracy and performance.

[0201] A "virtual store" refers to a store that is not dependent on a physical location and is operated on the Internet.

[0202] "Smart glasses" refers to a glasses-type device equipped with a display and sensors to provide visual information to the wearer.

[0203] "Real-time" refers to a state in which processing or response is carried out immediately the moment a specific operation or event occurs.

[0204] "Risk assessment" refers to the process of analyzing and evaluating the potential hazards and uncertainties associated with a particular action or scenario.

[0205] A "strategic proposal" refers to the presentation of a plan or method designed to achieve a specific objective.

[0206] "User" refers to a person such as a manager or entrepreneur who uses this system.

[0207] "Database" refers to a collection of data organized in a particular way and designed to be efficiently managed and searched.

[0208] "Terminal" refers to a device that is connected to a system or network and allows a user to input and output information.

[0209] The present invention relates to a system for providing support to a manager in making quick and appropriate management decisions in the operation of a virtual store. Specific embodiments for carrying out the present invention are described below.

[0210] System Configuration

[0211] server

[0212] The server includes the following functions:

[0213] 1. Data collection method: Use tools to collect information such as the manager's past statements, books, and video recordings. For this purpose, web scraping and APIs can be used. Voice recognition technology is applied to the video data to convert it into text data.

[0214] 2. Data cleaning method: The collected data is cleaned, noise is removed, and the format is standardized. The text data is tokenized and segmented into sentences, and then stored in a database.

[0215] 3. Generative AI model training method: Train a generative AI model (e.g., GPT-2) using the preprocessed data. Train the model using a deep learning framework (e.g., TensorFlow) and improve its performance through training and testing.

[0216] 4. Advice provision method: Generative AI models are used to generate appropriate advice for users' management inquiries. Scenario-based risk assessments and strategy proposals are provided in real time.

[0217] 5. Model Update Method: Regularly update the generative AI model with new data and feedback. Keep the model up to date by adding newly acquired data (e.g., interview or lecture data).

[0218] Terminal

[0219] The terminal includes the following features:

[0220] 1. Interface provision means: Provide an interface for virtual store operators to receive management decision advice in real time through smart glasses.

[0221] 2. Result display means: Displays the results of management decisions and advice. Depending on the scenario entered by the user, specific risk assessments and strategy proposals are displayed in text format.

[0222] User

[0223] The user does the following:

[0224] 1. Scenario input: Enter a specific business situation or scenario into the terminal. For example, enter a prompt such as, "Please tell us your risk assessment regarding the launch of a new product into the market."

[0225] 2. Refer to the results: Refer to the advice and evaluations provided by the generative AI model returned from the server and make actual management decisions.

[0226] Specific examples

[0227] Consider a case where a virtual store operator uses smart glasses to have a risk assessment performed by a generative AI model when introducing a new product. When the user types "Please tell me the risk assessment for the market launch of a new product" into the device interface, the server uses the generative AI model to generate an appropriate risk assessment and displays the result in real time on the smart glasses. This allows the operator to make an appropriate decision quickly on the spot.

[0228] As described above, the present invention provides virtual store operators with appropriate advice on business decisions in real time, thereby realizing more efficient operations.

[0229] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0230] Step 1:

[0231] The server collects data such as the manager's past statements, books, and video recordings. The input is data obtained from web scraping tools and APIs, and the output is text data or raw data before cleaning. This data is converted into text format using voice recognition technology (e.g., Google Speech-to-Text) or OCR.

[0232] Step 2:

[0233] The server cleans the collected data and standardizes its format. The input is the text data or raw data obtained in step 1, and the output is text data that has been denoised and integrated into a unified format. Specific operations include deleting duplicate data, removing unnecessary noise, tokenising (word splitting) and splitting sentences.

[0234] Step 3:

[0235] The server trains a generative AI model (e.g., GPT-2) using the preprocessed data. The input is the preprocessed text data, and the output is the trained generative AI model. Specifically, a deep learning framework (e.g., TensorFlow) is used to divide the text data into a training set and a test set, and to train the model and evaluate its performance.

[0236] Step 4:

[0237] The user inputs a specific business situation or scenario into the terminal interface. The input is a prompt, for example, "Please tell me your risk assessment regarding the launch of a new product into the market." The input prompt is then sent to the server.

[0238] Step 5:

[0239] The server uses a generative AI model to generate business decisions and advice based on the prompt text entered by the user. The input is the data from Step 4, which includes the prompt text and the generative AI model, and the output is the generated business decisions and advice text. The model generates optimal risk assessments and strategic proposals based on the input prompt text.

[0240] Step 6:

[0241] The terminal displays the management decisions and advice received from the server to the user. The input is the generated results sent from the server, and the output is text information displayed on a device such as smart glasses. Specifically, the user can check the advice in real time through the display of the smart glasses.

[0242] Step 7:

[0243] The server periodically updates the generative AI model using new data and user feedback. The input is the newly collected data or feedback, and the output is the updated generative AI model. This ensures that the model's accuracy and performance are always kept up to date.

[0244] 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.

[0245] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[0246] Data collection

[0247] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[0248] Data Preprocessing

[0249] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[0250] Model training

[0251] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[0252] Business decision simulation

[0253] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[0254] Management Consulting Platform

[0255] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[0256] Introducing the Emotion Engine

[0257] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module that analyzes voice data and an image analysis module that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[0258] Model emotional response

[0259] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[0260] Model Update

[0261] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[0262] Specific examples

[0263] Examples of data collection

[0264] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0265] Specific examples of business decision simulation

[0266] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0267] Examples of management consulting platforms

[0268] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0269] Examples of emotion engines

[0270] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[0271] The above is an embodiment of the present invention. By using this system, efficient and appropriate business decisions and business consultations can be made while inheriting the knowledge of the business owner, and more personalized advice can be provided that responds to the user's emotions.

[0272] The processing flow will be explained below.

[0273] Data collection

[0274] Step 1:

[0275] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[0276] Step 2:

[0277] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[0278] Step 3:

[0279] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[0280] Data Preprocessing

[0281] Step 1:

[0282] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[0283] Step 2:

[0284] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[0285] Step 3:

[0286] The server converts the cleaned data into a unified format and stores it in a database.

[0287] Model training

[0288] Step 1:

[0289] The server splits the preprocessed dataset into a training set and a test set.

[0290] Step 2:

[0291] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[0292] Step 3:

[0293] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[0294] Step 4:

[0295] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[0296] Business decision simulation

[0297] Step 1:

[0298] The terminal provides an interface to the simulation tool for internal users.

[0299] Step 2:

[0300] The user inputs a specific business scenario (e.g., new market entry, product development).

[0301] Step 3:

[0302] The server receives the input scenario and queries the generative AI model.

[0303] Step 4:

[0304] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[0305] Step 5:

[0306] The server returns the results to the terminal, where the user views the results.

[0307] Management Consulting Platform

[0308] Step 1:

[0309] The terminal provides an interface for the management consulting platform to external users.

[0310] Step 2:

[0311] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[0312] Step 3:

[0313] The server receives the input consultation content and executes a query against the generative AI model.

[0314] Step 4:

[0315] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[0316] Step 5:

[0317] The server sends advice and solutions to the terminal, which the user then confirms.

[0318] Introducing the Emotion Engine

[0319] Step 1:

[0320] The terminal activates a microphone for capturing the user's voice data and a camera for capturing facial expression data.

[0321] Step 2:

[0322] The server uses a voice analysis module to analyze the captured voice data and recognize the user's emotions (e.g., joy, sadness, anger, surprise).

[0323] Step 3:

[0324] The server uses an image analysis module to analyze the captured facial expression data and recognize the user's emotions.

[0325] Model emotional response

[0326] Step 1:

[0327] The server receives the user's emotional data analyzed by the emotion engine and determines the user's current emotional state.

[0328] Step 2:

[0329] The server adjusts the output of the generative AI model based on the determined emotion data. For example, if the user is nervous, the server adjusts the advice content to be gentler.

[0330] Step 3:

[0331] The generative AI model generates emotion-based tailored advice and business decisions and sends the results to a server.

[0332] Step 4:

[0333] The server sends the adjusted results to the terminal, and the user confirms them.

[0334] Model Update

[0335] Step 1:

[0336] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[0337] Step 2:

[0338] The server cleans the new data and standardizes the format.

[0339] Step 3:

[0340] The server uses the updated data to retrain the generative AI model and emotion engine.

[0341] Step 4:

[0342] The server evaluates the performance of the retrained model and releases an update.

[0343] The above are the specific processing steps of the program for the system that combines the emotion engine.

[0344] Example 2

[0345] 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."

[0346] Conventional management support systems have difficulty in fully utilizing the manager's past knowledge and experience, and in providing advice that reflects the user's feelings. As a result, they are unable to provide efficient and appropriate support for management decisions and management consultations, resulting in problems such as lower user satisfaction and success rates.

[0347] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for collecting the manager's past words and actions, books, video recordings, etc., a means for cleaning the collected data and standardizing the format, and a means for training the generative AI model using the preprocessed data. This makes it possible to provide efficient and appropriate management decisions and management consultations while inheriting the manager's knowledge.

[0348] "Past words and actions of management" refers to past words and actions made by management, as well as records of those actions.

[0349] "Books" refers to books written by business managers and books on management.

[0350] "Video recordings" refers to video or recorded media such as speeches or interviews in which the executive appears.

[0351] "Collection methods" refers to the technologies and methods used to obtain relevant data from the internet or physical media.

[0352] "Cleaning and formatting methods" refers to techniques and methods used to remove unnecessary information from collected data and convert it into a format suitable for analysis.

[0353] A "generative AI model" refers to an artificial intelligence model that is trained to reproduce the actions and behavior of managers.

[0354] "Preprocessed data" refers to data that has been cleaned and formatted and is used to train a model.

[0355] "Training means" refers to techniques and methods, including learning algorithms and computing resources, for improving the accuracy of artificial intelligence models.

[0356] "Business situations and scenarios" refer to specific business decision-making situations and strategies that users are considering.

[0357] "Means for generating business decisions" refers to technologies and methods for deriving optimal business decisions based on scenarios input by users using generative AI models.

[0358] "Advice using a generative AI model for business consultations" refers to advice and suggestions generated by AI in response to business-related questions and inquiries from users.

[0359] "Means of updating generative AI models with new data and feedback" refers to techniques and methods that incorporate new data collected and user feedback to improve the accuracy of the model and keep it up to date with the latest information.

[0360] An "emotion-recognizing emotion engine" refers to an artificial intelligence module that analyzes the user's emotional state from their voice and facial expressions.

[0361] "Means for adjusting output" refers to techniques or methods for changing the content or tone of the output of a generative AI model based on the user's emotional state.

[0362] "Speech recognition technology" refers to technology for converting voice data into text data.

[0363] "Optical Character Recognition (OCR)" refers to technology for extracting character data from images and scanned documents.

[0364] "Database" refers to a digital storage system that stores collected data in an organized manner and makes it easily searchable and accessible.

[0365] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[0366] Data collection

[0367] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the Internet. It then automatically converts video data into text using voice recognition technology (e.g., Google Speech-to-Text API), and digitizes book and article data using optical character recognition (OCR) technology (e.g., Tesseract OCR).

[0368] Data Preprocessing

[0369] The server then cleans and standardizes the collected data by removing duplicates and noise, tokenizing the text data (splitting words and sentences), and storing it in a consistent format in a database (e.g., MySQL, PostgreSQL).

[0370] Model training

[0371] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[0372] Business decision simulation

[0373] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[0374] Management Consulting Platform

[0375] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[0376] Introducing the Emotion Engine

[0377] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module (e.g., OpenSMILE) that analyzes voice data and an image analysis module (e.g., OpenCV, Dlib) that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[0378] Model emotional response

[0379] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[0380] Model Update

[0381] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[0382] Specific examples

[0383] Examples of data collection

[0384] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0385] Specific examples of business decision simulation

[0386] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0387] Examples of management consulting platforms

[0388] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0389] Examples of emotion engines

[0390] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[0391] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management, and it is possible to provide more personalized advice that responds to the user's emotions.

[0392] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0393] Program processing flow

[0394] Step 1: Data collection

[0395] The server collects the manager's past statements, actions, books, and video recordings from the internet. Specifically, it obtains the data using web scraping tools and APIs. Video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and book and article data is digitized using optical character recognition (OCR) technology (e.g., Tesseract OCR). The input is URLs and file paths related to the manager, and the output is a collection of text data.

[0396] Step 2: Data Preprocessing

[0397] The server cleans the collected data and standardizes its format. Specifically, it removes duplicate data and noise. It tokenizes (splits words) and sentences the text data and stores it in a consistent format in a database (e.g., MySQL, PostgreSQL). The input is the text data collected in step 1, and the output is organized data in a unified format.

[0398] Step 3: Model training

[0399] The server uses the preprocessed data to train a generative AI model. A deep learning framework (e.g., TensorFlow or PyTorch) is used to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning. The input is the preprocessed data from step 2, and the output is a trained generative AI model.

[0400] Step 4: Business decision simulation

[0401] The terminal provides an interface that allows internal users to input specific business situations and scenarios. Users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives the scenarios and simulates business decisions using a generative AI model. The results are sent back to the terminal, and the user refers to them to make the final business decision. The input is the business scenario, and the output is the simulation results.

[0402] Step 5: Management Consulting Platform

[0403] The terminal provides an interface for external users to receive management consultations. Users input specific questions about "fundraising strategies" and "recruitment methods." The server receives the input consultation content, generates appropriate advice using a generative AI model, and displays the output on the terminal. The user uses this advice as a reference when making management decisions. The input is the consultation content, and the output is the generated advice.

[0404] Step 6: Implementing the Emotion Engine

[0405] The device is equipped with a camera and microphone to capture the user's voice and facial expressions in real time. The server analyzes the voice data using a voice analysis module (e.g., OpenSMILE) and the facial expression data using an image analysis module (e.g., OpenCV, Dlib). The input is voice and facial expression data, and the output is the emotional data resulting from the analysis.

[0406] Step 7: Model emotional response

[0407] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, it will provide gentler advice, and if positive emotions are recognized, it will provide more proactive advice. The input is emotional data, and the output is the adjusted advice or judgment.

[0408] Step 8: Model Update

[0409] The server periodically updates the generative AI model based on new data and feedback. It adds newly collected interview and lecture data, integrates it with the existing dataset, and retrains the model. The input is the new data and feedback, and the output is an updated generative AI model.

[0410] The above are the specific processing steps of the system program. The operations performed at each step enable the user to make more efficient and appropriate business decisions and receive advice.

[0411] (Application example 2)

[0412] 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."

[0413] In today's business environment, it is difficult to pass on the knowledge and experience of managers to future generations, and there is a need for real-time management decisions and consultations that respond to users' emotions. Monitoring employee mental health is also an important issue, and it is difficult to provide effective countermeasures at the appropriate time. For this reason, there is a need to develop a system that can reproduce the words and actions of managers and provide real-time advice that responds to users' emotions.

[0414] 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 the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using preprocessed data, means for generating management decisions based on specific management situations and scenarios, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, an emotion recognition engine for recognizing the user's emotions, means for adjusting the output of the generative AI model based on emotion data using the emotion recognition engine, means for analyzing real-time data acquired from sensors, and means for providing appropriate responses and advice in real time using the generative AI model. This enables the provision of personalized advice that corresponds to the user's emotions and real-time mental health monitoring of employees while inheriting the manager's knowledge.

[0415] "Past words and actions of management" refers to records of statements, decisions, actions, etc. made by management in the past.

[0416] "Books" are books written by business managers or management reference works.

[0417] "Video recordings" refers to video data such as videos, interviews, and lectures featuring management.

[0418] "Means of collecting data" refers to methods of obtaining data, such as using web scraping tools, APIs, OCR technology, etc.

[0419] "Means for cleaning data and standardizing the format" refers to methods for removing duplication and noise from collected data, and for tokenizing and segmenting text data.

[0420] "Means for training a generative AI model using preprocessed data" refers to a method for training a generative AI model using cleaned data using a deep learning framework (e.g., TensorFlow or PyTorch).

[0421] A "means for generating business decisions" is a method for using a generative AI model to form business decisions based on specific business situations or scenarios.

[0422] The "means of providing advice" is a method of providing advice to a user's business consultation using a generative AI model.

[0423] A "generative AI model" is a model that uses deep learning technology to reproduce the words, actions, and thoughts of managers and generate management decisions and advice.

[0424] "Feedback-based updating" is a method of retraining a model based on new data and user feedback to improve its performance.

[0425] An "emotion recognition engine" is an engine that uses a voice analysis module and an image analysis module to recognize a user's emotions in real time.

[0426] "Means for adjusting output based on emotional data" refers to a method for adjusting the output results of a generative AI model based on emotional data analyzed by an emotion recognition engine.

[0427] "Means for analyzing real-time data" refers to methods for analyzing data obtained from security cameras and audio sensors in real time.

[0428] "Means for providing appropriate responses and advice in real time" refers to methods for providing appropriate responses and advice on the spot based on data analyzed using a generative AI model.

[0429] The present invention relates to a system that combines an emotion recognition engine and a generative AI model to provide management decisions based on the manager's past words and actions and experience, as well as real-time advice corresponding to the user's emotions. This system aims to make decisions in specific management situations and manage the mental health of employees. A specific embodiment of the present invention is described below.

[0430] Hardware and software used

[0431] 1. Hardware:

[0432] security cameras

[0433] Audio Sensor

[0434] microphone

[0435] GPU-equipped servers

[0436] 2. Software:

[0437] Python

[0438] OpenCV (image analysis library)

[0439] Keras (deep learning framework)

[0440] TensorFlow (deep learning framework)

[0441] SpeechRecognition library (audio analysis)

[0442] GPT-2 model (generative AI model)

[0443] Specific Embodiments of the System

[0444] 1. Data Collection

[0445] The server collects information such as past statements and actions of executives, books, and video recordings from online platforms and internal databases. The collected video data is converted into text using voice recognition technology, and books and articles are converted into digital data using OCR technology.

[0446] 2. Data Preprocessing

[0447] The server cleans the collected data, removes duplicate data and noise, tokenizes the text data, splits it into sentences, and stores it in a unified database.

[0448] 3. Model training

[0449] The server uses the preprocessed data to train a generative AI model. A deep learning framework (TensorFlow or PyTorch) is used to create the generative AI model. The model is divided into a training set and a test set, and its performance is improved through reinforcement learning and supervised learning.

[0450] 4. Emotion recognition

[0451] It analyzes real-time data (audio and video) of employees and users acquired through security cameras and audio sensors, and detects user emotions using an emotion recognition engine. It uses the SpeechRecognition library for audio analysis and OpenCV for image analysis.

[0452] 5. Tuning generative AI models based on emotion data

[0453] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion recognition engine. For example, if the user is feeling stressed, the generative AI model will be adjusted to output gentle and calming advice.

[0454] 6. Real-time advice

[0455] The server generates appropriate responses and advice based on real-time data analyzed using the generative AI model and displays them on the user's device, making it possible to monitor employee mental health and support management decision-making.

[0456] Specific examples

[0457] The server downloads videos of past executive speeches from an online platform, converts them into text using speech recognition technology, and stores them in a database after a cleaning process. An in-house project manager accesses a simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, which are then sent back to the terminal. A user using the terminal asks a question by voice, and the server uses a speech analysis module to analyze the user's emotions and determine that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[0458] Prompt Sentence Examples

[0459] "Generate examples of measures to take when an employee is under stress."

[0460] A system based on this format will utilize the knowledge of management, provide advice that responds to the user's emotions, and enable real-time mental health monitoring of employees.

[0461] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0462] Step 1:

[0463] The server collects information about the manager's past statements, actions, books, video recordings, etc. from online platforms and internal databases. Specifically, it acquires data using web scraping tools, APIs, and OCR technology. The inputs are the URL of the online platform and a database query, and the acquired data (text, audio, video) is output.

[0464] Step 2:

[0465] The server cleans the collected data by removing duplicates, removing noise, tokenizing text data, and segmenting sentences. The input is the collected raw data, and the output is cleaned, consistent data.

[0466] Step 3:

[0467] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework (TensorFlow or PyTorch) and performs reinforcement learning or supervised learning using a training set and a test set. It takes the preprocessed data as input and outputs a trained generative AI model.

[0468] Step 4:

[0469] A user accesses the business decision simulation tool using a terminal and inputs a specific business situation or scenario, such as "entering a new market" or "cost reduction strategy." The scenario information is input, and a simulation request is sent to the server.

[0470] Step 5:

[0471] The server uses a generative AI model to generate business decisions based on the specified scenario. Specifically, it analyzes the input scenario information and generates business decisions and strategic proposals. A scenario request is input, and the generated business decisions are output and sent to the terminal.

[0472] Step 6:

[0473] A user using a terminal accesses the business consulting platform and inputs a specific consultation content (e.g., "Fundraising strategy for business expansion"). With the consultation content as input, a consultation request is sent to the server.

[0474] Step 7:

[0475] The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. A consultation request is input, and the generated advice is output.

[0476] Step 8:

[0477] The device uses security cameras and audio sensors to capture real-time audio and video data to analyze the user's emotions. An emotion recognition engine is used to analyze the audio and facial expression data to identify the user's emotional state. Real-time data is input, and the emotion analysis results are output.

[0478] Step 9:

[0479] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion recognition engine. Specifically, if the user is nervous, it provides gentle and calm advice, and if positive emotions are recognized, it provides proactive advice. The input is the emotion analysis result, and the adjusted advice is output.

[0480] Step 10:

[0481] The server uses the generative AI model to provide appropriate responses and advice in real time, which are displayed on the device. This enables monitoring of employee mental health and supporting management decisions. The input is adjusted advice, and the final output is real-time responses and advice displayed on the user's device.

[0482] This realizes an efficient system that responds to the specific prompt statement mentioned above, "Please generate examples of measures to be taken when an employee is under stress."

[0483] 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.

[0484] 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.

[0485] 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.

[0486] [Second embodiment]

[0487] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0488] 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.

[0489] 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).

[0490] 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.

[0491] 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.

[0492] 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).

[0493] 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.

[0494] 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.

[0495] 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.

[0496] 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.

[0497] 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.

[0498] 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."

[0499] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[0500] Data collection

[0501] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[0502] Data Preprocessing

[0503] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[0504] Model training

[0505] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[0506] Business decision simulation

[0507] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[0508] Management Consulting Platform

[0509] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[0510] Model Update

[0511] The server periodically updates the generative AI model using new data and feedback, for example by collecting newly acquired executive interviews or speeches and adding them to the existing dataset, ensuring that the generative AI model always provides decisions based on the most up-to-date information.

[0512] Specific examples

[0513] Examples of data collection

[0514] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0515] Specific examples of business decision simulation

[0516] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0517] Examples of management consulting platforms

[0518] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0519] The above is an embodiment of the present invention. By using this system, it becomes possible to inherit the knowledge of the manager and to make efficient and appropriate management decisions and provide management consultations.

[0520] The processing flow will be explained below.

[0521] Data collection

[0522] Step 1:

[0523] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[0524] Step 2:

[0525] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[0526] Step 3:

[0527] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[0528] Data Preprocessing

[0529] Step 1:

[0530] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[0531] Step 2:

[0532] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[0533] Step 3:

[0534] The server converts the cleaned data into a unified format and stores it in a database.

[0535] Model training

[0536] Step 1:

[0537] The server splits the preprocessed dataset into a training set and a test set.

[0538] Step 2:

[0539] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[0540] Step 3:

[0541] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[0542] Step 4:

[0543] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[0544] Business decision simulation

[0545] Step 1:

[0546] The terminal provides an interface to the simulation tool for internal users.

[0547] Step 2:

[0548] The user inputs a specific business scenario (e.g., new market entry, product development).

[0549] Step 3:

[0550] The server receives the input scenario and queries the generative AI model.

[0551] Step 4:

[0552] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[0553] Step 5:

[0554] The server returns the results to the terminal, where the user views the results.

[0555] Management Consulting Platform

[0556] Step 1:

[0557] The terminal provides an interface for the management consulting platform to external users.

[0558] Step 2:

[0559] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[0560] Step 3:

[0561] The server receives the input consultation content and executes a query against the generative AI model.

[0562] Step 4:

[0563] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[0564] Step 5:

[0565] The server sends advice and solutions to the terminal, which the user then confirms.

[0566] Model Update

[0567] Step 1:

[0568] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[0569] Step 2:

[0570] The server cleans the new data and standardizes the format.

[0571] Step 3:

[0572] The server retrains the generative AI model with the updated data.

[0573] Step 4:

[0574] The server evaluates the performance of the retrained model and releases an update.

[0575] The above are the specific steps of the system program processing.

[0576] Example 1

[0577] 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."

[0578] There is a need for a system that can automatically and efficiently provide appropriate advice by systematizing management decisions that reflect the past words, actions, and thoughts of managers. However, current systems require cumbersome data collection and preprocessing, making it difficult to train effective generative AI models. Furthermore, users lack a means to instantly obtain appropriate management decisions and advice for specific scenarios. Furthermore, the process of updating the model based on new data and feedback is not automated, making it difficult to keep up with the latest management information.

[0579] 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.

[0580] In this invention, the server includes means for collecting the manager's past words and actions, literature, video recordings, etc., means for cleaning the collected data and standardizing the format, and means for training the generative AI model using the preprocessed data. This enables means for a user to input a specific management scenario and generate management decisions based on that scenario, means for providing advice using the generative AI model in response to a user's management consultation, and means for updating the generative AI model using new data and feedback.

[0581] A "management officer" is a person who is responsible for making important decisions in the operation of a company or organization.

[0582] "Behavior" refers collectively to speech and actions, and refers to words and actions shown in specific situations or on specific themes.

[0583] "Literature" is any kind of written material written to provide information, such as books, articles, or reports.

[0584] A "visual recording" is a collection of visual and audio information stored in the form of video, video recording or other media.

[0585] "Data collection" is the process of obtaining and gathering information needed for a specific purpose.

[0586] "Cleaning" is the process of removing errors and noise from collected data, organizing and normalizing the data.

[0587] "Unifying formats" is the process of converting data with different formats and structures into a consistent, standard format.

[0588] "Preprocessing" refers to data preparation performed before data analysis or model training.

[0589] A "generative AI model" is an algorithm that uses artificial intelligence techniques to learn specific patterns and characteristics and make inferences based on new data.

[0590] "Training" is the process of teaching a generative AI model using data to improve its performance.

[0591] A "business scenario" is a setting that represents assumptions and conditions regarding a particular business situation or strategy.

[0592] "Management decision-making" is the process by which managers make rational decisions regarding the operation of a company or organization.

[0593] "Management consultation" is the act of seeking professional advice on management issues and problems.

[0594] "Advice" is helpful advice or direction given on a specific problem or issue.

[0595] "Feedback" is evaluation or information provided to improve the performance of a system or model.

[0596] "Updating" is the process of adding new information or data to an existing system or model to improve its performance or accuracy.

[0597] A "database" is an organized collection of data that can be efficiently managed and searched.

[0598] A "user terminal" is a device or interface that a user can directly access and operate.

[0599] overview

[0600] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[0601] Data collection

[0602] The server first collects information on the manager's past statements and actions, documents, video recordings, etc. Tools used include web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). Furthermore, the video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and the document data is digitized using OCR technology (e.g., Tesseract).

[0603] Data Preprocessing

[0604] The server cleans the collected data and standardizes its format. The cleaning process involves removing duplicate data and removing noise. Tools used include NLP libraries (e.g., NLTK, spaCy). This data processing involves tokenizing the text data (splitting it into words) and storing it in a consistent format in the database.

[0605] Model training

[0606] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework such as TensorFlow or PyTorch. The dataset is divided into a training set and a test set, and the model's performance is evaluated and improved through cross-validation.

[0607] Business decision simulation

[0608] The terminal provides an interface that users can access and input specific business scenarios. For example, a web application may be provided that provides a form where users can input scenarios such as "entering a new market."

[0609] Users input scenarios through a simulation tool, and the server inputs prompt statements (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate business decisions.

[0610] The server sends the generated business decisions back to the terminal, allowing the user to view the results, for example by displaying the results on a dashboard so the user can view the details.

[0611] Management Consulting Platform

[0612] The terminal provides an interface for external users to receive management consultations. For example, it provides a chatbot and a consultation form, allowing users to input their consultation details, such as "fundraising strategies" or "recruitment methods."

[0613] The user inputs a specific business consultation, and the server generates advice using a generative AI model.

[0614] The server sends the generated advice back to the terminal, and the user makes management decisions based on that information. The advice is displayed in a chat window or in a report format.

[0615] Model Update

[0616] The server periodically updates the generative AI model with new data and feedback, for example by adding newly acquired executive interviews or speech data and updating the existing dataset.

[0617] The server retrains the model with the latest data to optimize its performance.

[0618] Specific examples

[0619] Examples of data collection

[0620] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0621] Specific examples of business decision simulation

[0622] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0623] Examples of management consulting platforms

[0624] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0625] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management.

[0626] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0627] Step 1: Data collection

[0628] The server collects data related to the manager. Input includes the manager's lecture videos, blog posts, interview articles, etc. from online platforms. Specific operations include using web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). These tools are used to automatically collect the necessary data and save it as a file. The output is raw data stored in local storage.

[0629] Step 2: Data Preprocessing

[0630] The server preprocesses the collected data. The input includes the raw data collected in step 1. Specifically, it converts voice data into text using speech recognition technology (e.g., Google Speech-to-Text API), and converts image and PDF data into digital text using OCR technology (e.g., Tesseract). It also performs noise removal, de-duplicate data removal, and tokenization using NLP libraries (e.g., NLTK, spaCy). The output is text data in a clean, unified format.

[0631] Step 3: Data storage

[0632] The server stores the preprocessed data in a database. The input includes the clean text data generated in step 2. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to store the text data in a structured format in the database. The output is the preprocessed text data stored in the database.

[0633] Step 4: Model training

[0634] The server uses the preprocessed data to train a generative AI model. The input includes the text data stored in the database in step 3. Specifically, a model is built using a deep learning framework (e.g., TensorFlow, PyTorch) and supervised learning and reinforcement learning are performed. The dataset is divided into a training set and a test set, and the model's performance is evaluated and optimized through cross-validation. The output is a generative AI model that can reproduce the words, actions, and thoughts of the manager.

[0635] Step 5: Business decision simulation

[0636] The terminal provides an interface for the user to input a business scenario. The input includes the business scenario (e.g., "Enter a new market") entered by the user. Specifically, it provides a web form or dashboard to allow the user to input the scenario. The server receives this scenario and inputs a prompt statement (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate a business decision. The output is the generated business decision, which is sent back to the terminal and displayed to the user.

[0637] Step 6: Management Consulting Platform

[0638] The terminal provides an interface for external users to input business advice. The input includes the specific advice entered by the user (e.g., "fundraising strategy"). Specific operations include providing a chatbot and a consultation form, allowing the user to input the details of the consultation. The server receives the consultation details and generates appropriate advice using a generative AI model. The output is the generated advice, which is sent back to the terminal and displayed to the user.

[0639] Step 7: Model Update

[0640] The server periodically updates the generative AI model using new data and feedback. The input includes newly collected managerial utterance data and feedback data. Specifically, it adds new data to the old dataset and retrains the model to optimize its performance. The output is an updated generative AI model, which can provide management decisions and advice based on the latest information.

[0641] (Application example 1)

[0642] 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."

[0643] In modern virtual store operations, managers must make a wide range of management decisions quickly and appropriately. However, relying on the manager's personal knowledge and experience can lead to biased decisions, resulting in insufficient risk assessment and strategic proposals. Furthermore, traditional management support systems have difficulty providing advice in real time, making them less effective in situations where immediate decisions are required on-site. This increases the burden on virtual store operators and hinders efficient operations, creating the challenge.

[0644] 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.

[0645] In this invention, the server includes means for collecting the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using the preprocessed data, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, means for providing management decision advice to the virtual store manager in real time via smart glasses, and means for generating risk assessments and strategic proposals based on scenarios specified by the user. This enables the virtual store manager to receive advice from the generative AI model in real time and make quick and appropriate management decisions.

[0646] A "management officer" is a person responsible for making decisions regarding the operation of a company or organization.

[0647] "Behavior" refers to all behavior, including verbal expressions and actions.

[0648] A "book" is a collection of bound documents that contain printed text and / or images and have a definite form.

[0649] "Visual recording" refers to data containing visual information in the form of video, film, etc.

[0650] "Collection methods" refer to methods and techniques for systematically gathering specific information or data.

[0651] "Cleaning" refers to the process of removing noise from data and arranging it into a unified format.

[0652] A "generative AI model" is an artificial intelligence model that learns using machine learning algorithms based on collected data.

[0653] "Training methods" refer to methods or techniques for training a generative AI model using a specific dataset.

[0654] "Business condition" refers to the current performance and environmental state of a company or organization.

[0655] A "scenario" refers to a hypothetical series of events or happenings based on specific circumstances or conditions.

[0656] "Advice delivery means" refers to the methods and technologies for communicating business decisions and proposals obtained by generative AI models to users.

[0657] "Update methods" refer to methods and techniques that incorporate new data and feedback into generative AI models to improve their accuracy and performance.

[0658] A "virtual store" refers to a store that is not dependent on a physical location and is operated on the Internet.

[0659] "Smart glasses" refers to a glasses-type device equipped with a display and sensors to provide visual information to the wearer.

[0660] "Real-time" refers to a state in which processing or response is carried out immediately the moment a specific operation or event occurs.

[0661] "Risk assessment" refers to the process of analyzing and evaluating the potential hazards and uncertainties associated with a particular action or scenario.

[0662] A "strategic proposal" refers to the presentation of a plan or method designed to achieve a specific objective.

[0663] "User" refers to a person such as a manager or entrepreneur who uses this system.

[0664] "Database" refers to a collection of data organized in a particular way and designed to be efficiently managed and searched.

[0665] "Terminal" refers to a device that is connected to a system or network and allows a user to input and output information.

[0666] The present invention relates to a system for providing support to a manager in making quick and appropriate management decisions in the operation of a virtual store. Specific embodiments for carrying out the present invention are described below.

[0667] System Configuration

[0668] server

[0669] The server includes the following functions:

[0670] 1. Data collection method: Use tools to collect information such as the manager's past statements, books, and video recordings. For this purpose, web scraping and APIs can be used. Voice recognition technology is applied to the video data to convert it into text data.

[0671] 2. Data cleaning method: The collected data is cleaned, noise is removed, and the format is standardized. The text data is tokenized and segmented into sentences, and then stored in a database.

[0672] 3. Generative AI model training method: Train a generative AI model (e.g., GPT-2) using the preprocessed data. Train the model using a deep learning framework (e.g., TensorFlow) and improve its performance through training and testing.

[0673] 4. Advice provision method: Generative AI models are used to generate appropriate advice for users' management inquiries. Scenario-based risk assessments and strategy proposals are provided in real time.

[0674] 5. Model Update Method: Regularly update the generative AI model with new data and feedback. Keep the model up to date by adding newly acquired data (e.g., interview or lecture data).

[0675] Terminal

[0676] The terminal includes the following features:

[0677] 1. Interface provision means: Provide an interface for virtual store operators to receive management decision advice in real time through smart glasses.

[0678] 2. Result display means: Displays the results of management decisions and advice. Depending on the scenario entered by the user, specific risk assessments and strategy proposals are displayed in text format.

[0679] User

[0680] The user does the following:

[0681] 1. Scenario input: Enter a specific business situation or scenario into the terminal. For example, enter a prompt such as, "Please tell us your risk assessment regarding the launch of a new product into the market."

[0682] 2. Refer to the results: Refer to the advice and evaluations provided by the generative AI model returned from the server and make actual management decisions.

[0683] Specific examples

[0684] Consider a case where a virtual store operator uses smart glasses to have a risk assessment performed by a generative AI model when introducing a new product. When the user types "Please tell me the risk assessment for the market launch of a new product" into the device interface, the server uses the generative AI model to generate an appropriate risk assessment and displays the result in real time on the smart glasses. This allows the operator to make an appropriate decision quickly on the spot.

[0685] As described above, the present invention provides virtual store operators with appropriate advice on business decisions in real time, thereby realizing more efficient operations.

[0686] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0687] Step 1:

[0688] The server collects data such as the manager's past statements, books, and video recordings. The input is data obtained from web scraping tools and APIs, and the output is text data or raw data before cleaning. This data is converted into text format using voice recognition technology (e.g., Google Speech-to-Text) or OCR.

[0689] Step 2:

[0690] The server cleans the collected data and standardizes its format. The input is the text data or raw data obtained in step 1, and the output is text data that has been denoised and integrated into a unified format. Specific operations include deleting duplicate data, removing unnecessary noise, tokenising (word splitting) and splitting sentences.

[0691] Step 3:

[0692] The server trains a generative AI model (e.g., GPT-2) using the preprocessed data. The input is the preprocessed text data, and the output is the trained generative AI model. Specifically, a deep learning framework (e.g., TensorFlow) is used to divide the text data into a training set and a test set, and to train the model and evaluate its performance.

[0693] Step 4:

[0694] The user inputs a specific business situation or scenario into the terminal interface. The input is a prompt, for example, "Please tell me your risk assessment regarding the launch of a new product into the market." The input prompt is then sent to the server.

[0695] Step 5:

[0696] The server uses a generative AI model to generate business decisions and advice based on the prompt text entered by the user. The input is the data from Step 4, which includes the prompt text and the generative AI model, and the output is the generated business decisions and advice text. The model generates optimal risk assessments and strategic proposals based on the input prompt text.

[0697] Step 6:

[0698] The terminal displays the management decisions and advice received from the server to the user. The input is the generated results sent from the server, and the output is text information displayed on a device such as smart glasses. Specifically, the user can check the advice in real time through the display of the smart glasses.

[0699] Step 7:

[0700] The server periodically updates the generative AI model using new data and user feedback. The input is the newly collected data or feedback, and the output is the updated generative AI model. This ensures that the model's accuracy and performance are always kept up to date.

[0701] 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.

[0702] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[0703] Data collection

[0704] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[0705] Data Preprocessing

[0706] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[0707] Model training

[0708] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[0709] Business decision simulation

[0710] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[0711] Management Consulting Platform

[0712] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[0713] Introducing the Emotion Engine

[0714] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module that analyzes voice data and an image analysis module that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[0715] Model emotional response

[0716] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[0717] Model Update

[0718] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[0719] Specific examples

[0720] Examples of data collection

[0721] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0722] Specific examples of business decision simulation

[0723] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0724] Examples of management consulting platforms

[0725] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0726] Examples of emotion engines

[0727] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[0728] The above is an embodiment of the present invention. By using this system, efficient and appropriate business decisions and business consultations can be made while inheriting the knowledge of the business owner, and more personalized advice can be provided that responds to the user's emotions.

[0729] The processing flow will be explained below.

[0730] Data collection

[0731] Step 1:

[0732] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[0733] Step 2:

[0734] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[0735] Step 3:

[0736] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[0737] Data Preprocessing

[0738] Step 1:

[0739] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[0740] Step 2:

[0741] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[0742] Step 3:

[0743] The server converts the cleaned data into a unified format and stores it in a database.

[0744] Model training

[0745] Step 1:

[0746] The server splits the preprocessed dataset into a training set and a test set.

[0747] Step 2:

[0748] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[0749] Step 3:

[0750] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[0751] Step 4:

[0752] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[0753] Business decision simulation

[0754] Step 1:

[0755] The terminal provides an interface to the simulation tool for internal users.

[0756] Step 2:

[0757] The user inputs a specific business scenario (e.g., new market entry, product development).

[0758] Step 3:

[0759] The server receives the input scenario and queries the generative AI model.

[0760] Step 4:

[0761] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[0762] Step 5:

[0763] The server returns the results to the terminal, where the user views the results.

[0764] Management Consulting Platform

[0765] Step 1:

[0766] The terminal provides an interface for the management consulting platform to external users.

[0767] Step 2:

[0768] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[0769] Step 3:

[0770] The server receives the input consultation content and executes a query against the generative AI model.

[0771] Step 4:

[0772] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[0773] Step 5:

[0774] The server sends advice and solutions to the terminal, which the user then confirms.

[0775] Introducing the Emotion Engine

[0776] Step 1:

[0777] The terminal activates a microphone for capturing the user's voice data and a camera for capturing facial expression data.

[0778] Step 2:

[0779] The server uses a voice analysis module to analyze the captured voice data and recognize the user's emotions (e.g., joy, sadness, anger, surprise).

[0780] Step 3:

[0781] The server uses an image analysis module to analyze the captured facial expression data and recognize the user's emotions.

[0782] Model emotional response

[0783] Step 1:

[0784] The server receives the user's emotional data analyzed by the emotion engine and determines the user's current emotional state.

[0785] Step 2:

[0786] The server adjusts the output of the generative AI model based on the determined emotion data. For example, if the user is nervous, the server adjusts the advice content to be gentler.

[0787] Step 3:

[0788] The generative AI model generates emotion-based tailored advice and business decisions and sends the results to a server.

[0789] Step 4:

[0790] The server sends the adjusted results to the terminal, and the user confirms them.

[0791] Model Update

[0792] Step 1:

[0793] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[0794] Step 2:

[0795] The server cleans the new data and standardizes the format.

[0796] Step 3:

[0797] The server uses the updated data to retrain the generative AI model and emotion engine.

[0798] Step 4:

[0799] The server evaluates the performance of the retrained model and releases an update.

[0800] The above are the specific processing steps of the program for the system that combines the emotion engine.

[0801] Example 2

[0802] 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."

[0803] Conventional management support systems have difficulty in fully utilizing the manager's past knowledge and experience, and in providing advice that reflects the user's feelings. As a result, they are unable to provide efficient and appropriate support for management decisions and management consultations, resulting in problems such as lower user satisfaction and success rates.

[0804] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for collecting the manager's past words and actions, books, video recordings, etc., a means for cleaning the collected data and standardizing the format, and a means for training the generative AI model using the preprocessed data. This makes it possible to provide efficient and appropriate management decisions and management consultations while inheriting the manager's knowledge.

[0805] "Past words and actions of management" refers to past words and actions made by management, as well as records of those actions.

[0806] "Books" refers to books written by business managers and books on management.

[0807] "Video recordings" refers to video or recorded media such as speeches or interviews in which the executive appears.

[0808] "Collection methods" refers to the technologies and methods used to obtain relevant data from the internet or physical media.

[0809] "Cleaning and formatting methods" refers to techniques and methods used to remove unnecessary information from collected data and convert it into a format suitable for analysis.

[0810] A "generative AI model" refers to an artificial intelligence model that is trained to reproduce the actions and behavior of managers.

[0811] "Preprocessed data" refers to data that has been cleaned and formatted and is used to train a model.

[0812] "Training means" refers to techniques and methods, including learning algorithms and computing resources, for improving the accuracy of artificial intelligence models.

[0813] "Business situations and scenarios" refer to specific business decision-making situations and strategies that users are considering.

[0814] "Means for generating business decisions" refers to technologies and methods for deriving optimal business decisions based on scenarios input by users using generative AI models.

[0815] "Advice using a generative AI model for business consultations" refers to advice and suggestions generated by AI in response to business-related questions and inquiries from users.

[0816] "Means of updating generative AI models with new data and feedback" refers to techniques and methods that incorporate new data collected and user feedback to improve the accuracy of the model and keep it up to date with the latest information.

[0817] An "emotion-recognizing emotion engine" refers to an artificial intelligence module that analyzes the user's emotional state from their voice and facial expressions.

[0818] "Means for adjusting output" refers to techniques or methods for changing the content or tone of the output of a generative AI model based on the user's emotional state.

[0819] "Speech recognition technology" refers to technology for converting voice data into text data.

[0820] "Optical Character Recognition (OCR)" refers to technology for extracting character data from images and scanned documents.

[0821] "Database" refers to a digital storage system that stores collected data in an organized manner and makes it easily searchable and accessible.

[0822] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[0823] Data collection

[0824] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the Internet. It then automatically converts video data into text using voice recognition technology (e.g., Google Speech-to-Text API), and digitizes book and article data using optical character recognition (OCR) technology (e.g., Tesseract OCR).

[0825] Data Preprocessing

[0826] The server then cleans and standardizes the collected data by removing duplicates and noise, tokenizing the text data (splitting words and sentences), and storing it in a consistent format in a database (e.g., MySQL, PostgreSQL).

[0827] Model training

[0828] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[0829] Business decision simulation

[0830] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[0831] Management Consulting Platform

[0832] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[0833] Introducing the Emotion Engine

[0834] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module (e.g., OpenSMILE) that analyzes voice data and an image analysis module (e.g., OpenCV, Dlib) that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[0835] Model emotional response

[0836] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[0837] Model Update

[0838] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[0839] Specific examples

[0840] Examples of data collection

[0841] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0842] Specific examples of business decision simulation

[0843] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0844] Examples of management consulting platforms

[0845] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0846] Examples of emotion engines

[0847] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[0848] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management, and it is possible to provide more personalized advice that responds to the user's emotions.

[0849] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0850] Program processing flow

[0851] Step 1: Data collection

[0852] The server collects the manager's past statements, actions, books, and video recordings from the internet. Specifically, it obtains the data using web scraping tools and APIs. Video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and book and article data is digitized using optical character recognition (OCR) technology (e.g., Tesseract OCR). The input is URLs and file paths related to the manager, and the output is a collection of text data.

[0853] Step 2: Data Preprocessing

[0854] The server cleans the collected data and standardizes its format. Specifically, it removes duplicate data and noise. It tokenizes (splits words) and sentences the text data and stores it in a consistent format in a database (e.g., MySQL, PostgreSQL). The input is the text data collected in step 1, and the output is organized data in a unified format.

[0855] Step 3: Model training

[0856] The server uses the preprocessed data to train a generative AI model. A deep learning framework (e.g., TensorFlow or PyTorch) is used to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning. The input is the preprocessed data from step 2, and the output is a trained generative AI model.

[0857] Step 4: Business decision simulation

[0858] The terminal provides an interface that allows internal users to input specific business situations and scenarios. Users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives the scenarios and simulates business decisions using a generative AI model. The results are sent back to the terminal, and the user refers to them to make the final business decision. The input is the business scenario, and the output is the simulation results.

[0859] Step 5: Management Consulting Platform

[0860] The terminal provides an interface for external users to receive management consultations. Users input specific questions about "fundraising strategies" and "recruitment methods." The server receives the input consultation content, generates appropriate advice using a generative AI model, and displays the output on the terminal. The user uses this advice as a reference when making management decisions. The input is the consultation content, and the output is the generated advice.

[0861] Step 6: Implementing the Emotion Engine

[0862] The device is equipped with a camera and microphone to capture the user's voice and facial expressions in real time. The server analyzes the voice data using a voice analysis module (e.g., OpenSMILE) and the facial expression data using an image analysis module (e.g., OpenCV, Dlib). The input is voice and facial expression data, and the output is the emotional data resulting from the analysis.

[0863] Step 7: Model emotional response

[0864] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, it will provide gentler advice, and if positive emotions are recognized, it will provide more proactive advice. The input is emotional data, and the output is the adjusted advice or judgment.

[0865] Step 8: Model Update

[0866] The server periodically updates the generative AI model based on new data and feedback. It adds newly collected interview and lecture data, integrates it with the existing dataset, and retrains the model. The input is the new data and feedback, and the output is an updated generative AI model.

[0867] The above are the specific processing steps of the system program. The operations performed at each step enable the user to make more efficient and appropriate business decisions and receive advice.

[0868] (Application example 2)

[0869] 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."

[0870] In today's business environment, it is difficult to pass on the knowledge and experience of managers to future generations, and there is a need for real-time management decisions and consultations that respond to users' emotions. Monitoring employee mental health is also an important issue, and it is difficult to provide effective countermeasures at the appropriate time. For this reason, there is a need to develop a system that can reproduce the words and actions of managers and provide real-time advice that responds to users' emotions.

[0871] 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 the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using preprocessed data, means for generating management decisions based on specific management situations and scenarios, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, an emotion recognition engine for recognizing the user's emotions, means for adjusting the output of the generative AI model based on emotion data using the emotion recognition engine, means for analyzing real-time data acquired from sensors, and means for providing appropriate responses and advice in real time using the generative AI model. This enables the provision of personalized advice that corresponds to the user's emotions and real-time mental health monitoring of employees while inheriting the manager's knowledge.

[0872] "Past words and actions of management" refers to records of statements, decisions, actions, etc. made by management in the past.

[0873] "Books" are books written by business managers or management reference works.

[0874] "Video recordings" refers to video data such as videos, interviews, and lectures featuring management.

[0875] "Means of collecting data" refers to methods of obtaining data, such as using web scraping tools, APIs, OCR technology, etc.

[0876] "Means for cleaning data and standardizing the format" refers to methods for removing duplication and noise from collected data, and for tokenizing and segmenting text data.

[0877] "Means for training a generative AI model using preprocessed data" refers to a method for training a generative AI model using cleaned data using a deep learning framework (e.g., TensorFlow or PyTorch).

[0878] A "means for generating business decisions" is a method for using a generative AI model to form business decisions based on specific business situations or scenarios.

[0879] The "means of providing advice" is a method of providing advice to a user's business consultation using a generative AI model.

[0880] A "generative AI model" is a model that uses deep learning technology to reproduce the words, actions, and thoughts of managers and generate management decisions and advice.

[0881] "Feedback-based updating" is a method of retraining a model based on new data and user feedback to improve its performance.

[0882] An "emotion recognition engine" is an engine that uses a voice analysis module and an image analysis module to recognize a user's emotions in real time.

[0883] "Means for adjusting output based on emotional data" refers to a method for adjusting the output results of a generative AI model based on emotional data analyzed by an emotion recognition engine.

[0884] "Means for analyzing real-time data" refers to methods for analyzing data obtained from security cameras and audio sensors in real time.

[0885] "Means for providing appropriate responses and advice in real time" refers to methods for providing appropriate responses and advice on the spot based on data analyzed using a generative AI model.

[0886] The present invention relates to a system that combines an emotion recognition engine and a generative AI model to provide management decisions based on the manager's past words and actions and experience, as well as real-time advice corresponding to the user's emotions. This system aims to make decisions in specific management situations and manage the mental health of employees. A specific embodiment of the present invention is described below.

[0887] Hardware and software used

[0888] 1. Hardware:

[0889] security cameras

[0890] Audio Sensor

[0891] microphone

[0892] GPU-equipped servers

[0893] 2. Software:

[0894] Python

[0895] OpenCV (image analysis library)

[0896] Keras (deep learning framework)

[0897] TensorFlow (deep learning framework)

[0898] SpeechRecognition library (audio analysis)

[0899] GPT-2 model (generative AI model)

[0900] Specific Embodiments of the System

[0901] 1. Data Collection

[0902] The server collects information such as past statements and actions of executives, books, and video recordings from online platforms and internal databases. The collected video data is converted into text using voice recognition technology, and books and articles are converted into digital data using OCR technology.

[0903] 2. Data Preprocessing

[0904] The server cleans the collected data, removes duplicate data and noise, tokenizes the text data, splits it into sentences, and stores it in a unified database.

[0905] 3. Model training

[0906] The server uses the preprocessed data to train a generative AI model. A deep learning framework (TensorFlow or PyTorch) is used to create the generative AI model. The model is divided into a training set and a test set, and its performance is improved through reinforcement learning and supervised learning.

[0907] 4. Emotion recognition

[0908] It analyzes real-time data (audio and video) of employees and users acquired through security cameras and audio sensors, and detects user emotions using an emotion recognition engine. It uses the SpeechRecognition library for audio analysis and OpenCV for image analysis.

[0909] 5. Tuning generative AI models based on emotion data

[0910] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion recognition engine. For example, if the user is feeling stressed, the generative AI model will be adjusted to output gentle and calming advice.

[0911] 6. Real-time advice

[0912] The server generates appropriate responses and advice based on real-time data analyzed using the generative AI model and displays them on the user's device, making it possible to monitor employee mental health and support management decision-making.

[0913] Specific examples

[0914] The server downloads videos of past executive speeches from an online platform, converts them into text using speech recognition technology, and stores them in a database after a cleaning process. An in-house project manager accesses a simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, which are then sent back to the terminal. A user using the terminal asks a question by voice, and the server uses a speech analysis module to analyze the user's emotions and determine that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[0915] Prompt Sentence Examples

[0916] "Generate examples of measures to take when an employee is under stress."

[0917] A system based on this format will utilize the knowledge of management, provide advice that responds to the user's emotions, and enable real-time mental health monitoring of employees.

[0918] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0919] Step 1:

[0920] The server collects information about the manager's past statements, actions, books, video recordings, etc. from online platforms and internal databases. Specifically, it acquires data using web scraping tools, APIs, and OCR technology. The inputs are the URL of the online platform and a database query, and the acquired data (text, audio, video) is output.

[0921] Step 2:

[0922] The server cleans the collected data by removing duplicates, removing noise, tokenizing text data, and segmenting sentences. The input is the collected raw data, and the output is cleaned, consistent data.

[0923] Step 3:

[0924] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework (TensorFlow or PyTorch) and performs reinforcement learning or supervised learning using a training set and a test set. It takes the preprocessed data as input and outputs a trained generative AI model.

[0925] Step 4:

[0926] A user accesses the business decision simulation tool using a terminal and inputs a specific business situation or scenario, such as "entering a new market" or "cost reduction strategy." The scenario information is input, and a simulation request is sent to the server.

[0927] Step 5:

[0928] The server uses a generative AI model to generate business decisions based on the specified scenario. Specifically, it analyzes the input scenario information and generates business decisions and strategic proposals. A scenario request is input, and the generated business decisions are output and sent to the terminal.

[0929] Step 6:

[0930] A user using a terminal accesses the business consulting platform and inputs a specific consultation content (e.g., "Fundraising strategy for business expansion"). With the consultation content as input, a consultation request is sent to the server.

[0931] Step 7:

[0932] The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. A consultation request is input, and the generated advice is output.

[0933] Step 8:

[0934] The device uses security cameras and audio sensors to capture real-time audio and video data to analyze the user's emotions. An emotion recognition engine is used to analyze the audio and facial expression data to identify the user's emotional state. Real-time data is input, and the emotion analysis results are output.

[0935] Step 9:

[0936] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion recognition engine. Specifically, if the user is nervous, it provides gentle and calm advice, and if positive emotions are recognized, it provides proactive advice. The input is the emotion analysis result, and the adjusted advice is output.

[0937] Step 10:

[0938] The server uses the generative AI model to provide appropriate responses and advice in real time, which are displayed on the device. This enables monitoring of employee mental health and supporting management decisions. The input is adjusted advice, and the final output is real-time responses and advice displayed on the user's device.

[0939] This realizes an efficient system that responds to the specific prompt statement mentioned above, "Please generate examples of measures to be taken when an employee is under stress."

[0940] 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.

[0941] 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.

[0942] 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.

[0943] [Third embodiment]

[0944] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0945] 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.

[0946] 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).

[0947] 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.

[0948] 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.

[0949] 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).

[0950] 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.

[0951] 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.

[0952] 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.

[0953] 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.

[0954] 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.

[0955] 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."

[0956] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[0957] Data collection

[0958] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[0959] Data Preprocessing

[0960] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[0961] Model training

[0962] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[0963] Business decision simulation

[0964] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[0965] Management Consulting Platform

[0966] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[0967] Model Update

[0968] The server periodically updates the generative AI model using new data and feedback, for example by collecting newly acquired executive interviews or speeches and adding them to the existing dataset, ensuring that the generative AI model always provides decisions based on the most up-to-date information.

[0969] Specific examples

[0970] Examples of data collection

[0971] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[0972] Specific examples of business decision simulation

[0973] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[0974] Examples of management consulting platforms

[0975] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[0976] The above is an embodiment of the present invention. By using this system, it becomes possible to inherit the knowledge of the manager and to make efficient and appropriate management decisions and provide management consultations.

[0977] The processing flow will be explained below.

[0978] Data collection

[0979] Step 1:

[0980] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[0981] Step 2:

[0982] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[0983] Step 3:

[0984] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[0985] Data Preprocessing

[0986] Step 1:

[0987] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[0988] Step 2:

[0989] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[0990] Step 3:

[0991] The server converts the cleaned data into a unified format and stores it in a database.

[0992] Model training

[0993] Step 1:

[0994] The server splits the preprocessed dataset into a training set and a test set.

[0995] Step 2:

[0996] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[0997] Step 3:

[0998] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[0999] Step 4:

[1000] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[1001] Business decision simulation

[1002] Step 1:

[1003] The terminal provides an interface to the simulation tool for internal users.

[1004] Step 2:

[1005] The user inputs a specific business scenario (e.g., new market entry, product development).

[1006] Step 3:

[1007] The server receives the input scenario and queries the generative AI model.

[1008] Step 4:

[1009] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[1010] Step 5:

[1011] The server returns the results to the terminal, where the user views the results.

[1012] Management Consulting Platform

[1013] Step 1:

[1014] The terminal provides an interface for the management consulting platform to external users.

[1015] Step 2:

[1016] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[1017] Step 3:

[1018] The server receives the input consultation content and executes a query against the generative AI model.

[1019] Step 4:

[1020] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[1021] Step 5:

[1022] The server sends advice and solutions to the terminal, which the user then confirms.

[1023] Model Update

[1024] Step 1:

[1025] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[1026] Step 2:

[1027] The server cleans the new data and standardizes the format.

[1028] Step 3:

[1029] The server retrains the generative AI model with the updated data.

[1030] Step 4:

[1031] The server evaluates the performance of the retrained model and releases an update.

[1032] The above are the specific steps of the system program processing.

[1033] Example 1

[1034] 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."

[1035] There is a need for a system that can automatically and efficiently provide appropriate advice by systematizing management decisions that reflect the past words, actions, and thoughts of managers. However, current systems require cumbersome data collection and preprocessing, making it difficult to train effective generative AI models. Furthermore, users lack a means to instantly obtain appropriate management decisions and advice for specific scenarios. Furthermore, the process of updating the model based on new data and feedback is not automated, making it difficult to keep up with the latest management information.

[1036] 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.

[1037] In this invention, the server includes means for collecting the manager's past words and actions, literature, video recordings, etc., means for cleaning the collected data and standardizing the format, and means for training the generative AI model using the preprocessed data. This enables means for a user to input a specific management scenario and generate management decisions based on that scenario, means for providing advice using the generative AI model in response to a user's management consultation, and means for updating the generative AI model using new data and feedback.

[1038] A "management officer" is a person who is responsible for making important decisions in the operation of a company or organization.

[1039] "Behavior" refers collectively to speech and actions, and refers to words and actions shown in specific situations or on specific themes.

[1040] "Literature" is any kind of written material written to provide information, such as books, articles, or reports.

[1041] A "visual recording" is a collection of visual and audio information stored in the form of video, video recording or other media.

[1042] "Data collection" is the process of obtaining and gathering information needed for a specific purpose.

[1043] "Cleaning" is the process of removing errors and noise from collected data, organizing and normalizing the data.

[1044] "Unifying formats" is the process of converting data with different formats and structures into a consistent, standard format.

[1045] "Preprocessing" refers to data preparation performed before data analysis or model training.

[1046] A "generative AI model" is an algorithm that uses artificial intelligence techniques to learn specific patterns and characteristics and make inferences based on new data.

[1047] "Training" is the process of teaching a generative AI model using data to improve its performance.

[1048] A "business scenario" is a setting that represents assumptions and conditions regarding a particular business situation or strategy.

[1049] "Management decision-making" is the process by which managers make rational decisions regarding the operation of a company or organization.

[1050] "Management consultation" is the act of seeking professional advice on management issues and problems.

[1051] "Advice" is helpful advice or direction given on a specific problem or issue.

[1052] "Feedback" is evaluation or information provided to improve the performance of a system or model.

[1053] "Updating" is the process of adding new information or data to an existing system or model to improve its performance or accuracy.

[1054] A "database" is an organized collection of data that can be efficiently managed and searched.

[1055] A "user terminal" is a device or interface that a user can directly access and operate.

[1056] overview

[1057] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[1058] Data collection

[1059] The server first collects information on the manager's past statements and actions, documents, video recordings, etc. Tools used include web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). Furthermore, the video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and the document data is digitized using OCR technology (e.g., Tesseract).

[1060] Data Preprocessing

[1061] The server cleans the collected data and standardizes its format. The cleaning process involves removing duplicate data and removing noise. Tools used include NLP libraries (e.g., NLTK, spaCy). This data processing involves tokenizing the text data (splitting it into words) and storing it in a consistent format in the database.

[1062] Model training

[1063] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework such as TensorFlow or PyTorch. The dataset is divided into a training set and a test set, and the model's performance is evaluated and improved through cross-validation.

[1064] Business decision simulation

[1065] The terminal provides an interface that users can access and input specific business scenarios. For example, a web application may be provided that provides a form where users can input scenarios such as "entering a new market."

[1066] Users input scenarios through a simulation tool, and the server inputs prompt statements (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate business decisions.

[1067] The server sends the generated business decisions back to the terminal, allowing the user to view the results, for example by displaying the results on a dashboard so the user can view the details.

[1068] Management Consulting Platform

[1069] The terminal provides an interface for external users to receive management consultations. For example, it provides a chatbot and a consultation form, allowing users to input their consultation details, such as "fundraising strategies" or "recruitment methods."

[1070] The user inputs a specific business consultation, and the server generates advice using a generative AI model.

[1071] The server sends the generated advice back to the terminal, and the user makes management decisions based on that information. The advice is displayed in a chat window or in a report format.

[1072] Model Update

[1073] The server periodically updates the generative AI model with new data and feedback, for example by adding newly acquired executive interviews or speech data and updating the existing dataset.

[1074] The server retrains the model with the latest data to optimize its performance.

[1075] Specific examples

[1076] Examples of data collection

[1077] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[1078] Specific examples of business decision simulation

[1079] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[1080] Examples of management consulting platforms

[1081] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[1082] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management.

[1083] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1084] Step 1: Data collection

[1085] The server collects data related to the manager. Input includes the manager's lecture videos, blog posts, interview articles, etc. from online platforms. Specific operations include using web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). These tools are used to automatically collect the necessary data and save it as a file. The output is raw data stored in local storage.

[1086] Step 2: Data Preprocessing

[1087] The server preprocesses the collected data. The input includes the raw data collected in step 1. Specifically, it converts voice data into text using speech recognition technology (e.g., Google Speech-to-Text API), and converts image and PDF data into digital text using OCR technology (e.g., Tesseract). It also performs noise removal, de-duplicate data removal, and tokenization using NLP libraries (e.g., NLTK, spaCy). The output is text data in a clean, unified format.

[1088] Step 3: Data storage

[1089] The server stores the preprocessed data in a database. The input includes the clean text data generated in step 2. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to store the text data in a structured format in the database. The output is the preprocessed text data stored in the database.

[1090] Step 4: Model training

[1091] The server uses the preprocessed data to train a generative AI model. The input includes the text data stored in the database in step 3. Specifically, a model is built using a deep learning framework (e.g., TensorFlow, PyTorch) and supervised learning and reinforcement learning are performed. The dataset is divided into a training set and a test set, and the model's performance is evaluated and optimized through cross-validation. The output is a generative AI model that can reproduce the words, actions, and thoughts of the manager.

[1092] Step 5: Business decision simulation

[1093] The terminal provides an interface for the user to input a business scenario. The input includes the business scenario (e.g., "Enter a new market") entered by the user. Specifically, it provides a web form or dashboard to allow the user to input the scenario. The server receives this scenario and inputs a prompt statement (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate a business decision. The output is the generated business decision, which is sent back to the terminal and displayed to the user.

[1094] Step 6: Management Consulting Platform

[1095] The terminal provides an interface for external users to input business advice. The input includes the specific advice entered by the user (e.g., "fundraising strategy"). Specific operations include providing a chatbot and a consultation form, allowing the user to input the details of the consultation. The server receives the consultation details and generates appropriate advice using a generative AI model. The output is the generated advice, which is sent back to the terminal and displayed to the user.

[1096] Step 7: Model Update

[1097] The server periodically updates the generative AI model using new data and feedback. The input includes newly collected managerial utterance data and feedback data. Specifically, it adds new data to the old dataset and retrains the model to optimize its performance. The output is an updated generative AI model, which can provide management decisions and advice based on the latest information.

[1098] (Application example 1)

[1099] 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."

[1100] In modern virtual store operations, managers must make a wide range of management decisions quickly and appropriately. However, relying on the manager's personal knowledge and experience can lead to biased decisions, resulting in insufficient risk assessment and strategic proposals. Furthermore, traditional management support systems have difficulty providing advice in real time, making them less effective in situations where immediate decisions are required on-site. This increases the burden on virtual store operators and hinders efficient operations, creating the challenge.

[1101] 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.

[1102] In this invention, the server includes means for collecting the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using the preprocessed data, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, means for providing management decision advice to the virtual store manager in real time via smart glasses, and means for generating risk assessments and strategic proposals based on scenarios specified by the user. This enables the virtual store manager to receive advice from the generative AI model in real time and make quick and appropriate management decisions.

[1103] A "management officer" is a person responsible for making decisions regarding the operation of a company or organization.

[1104] "Behavior" refers to all behavior, including verbal expressions and actions.

[1105] A "book" is a collection of bound documents that contain printed text and / or images and have a definite form.

[1106] "Visual recording" refers to data containing visual information in the form of video, film, etc.

[1107] "Collection methods" refer to methods and techniques for systematically gathering specific information or data.

[1108] "Cleaning" refers to the process of removing noise from data and arranging it into a unified format.

[1109] A "generative AI model" is an artificial intelligence model that learns using machine learning algorithms based on collected data.

[1110] "Training methods" refer to methods or techniques for training a generative AI model using a specific dataset.

[1111] "Business condition" refers to the current performance and environmental state of a company or organization.

[1112] A "scenario" refers to a hypothetical series of events or happenings based on specific circumstances or conditions.

[1113] "Advice delivery means" refers to the methods and technologies for communicating business decisions and proposals obtained by generative AI models to users.

[1114] "Update methods" refer to methods and techniques that incorporate new data and feedback into generative AI models to improve their accuracy and performance.

[1115] A "virtual store" refers to a store that is not dependent on a physical location and is operated on the Internet.

[1116] "Smart glasses" refers to a glasses-type device equipped with a display and sensors to provide visual information to the wearer.

[1117] "Real-time" refers to a state in which processing or response is carried out immediately the moment a specific operation or event occurs.

[1118] "Risk assessment" refers to the process of analyzing and evaluating the potential hazards and uncertainties associated with a particular action or scenario.

[1119] A "strategic proposal" refers to the presentation of a plan or method designed to achieve a specific objective.

[1120] "User" refers to a person such as a manager or entrepreneur who uses this system.

[1121] "Database" refers to a collection of data organized in a particular way and designed to be efficiently managed and searched.

[1122] "Terminal" refers to a device that is connected to a system or network and allows a user to input and output information.

[1123] The present invention relates to a system for providing support to a manager in making quick and appropriate management decisions in the operation of a virtual store. Specific embodiments for carrying out the present invention are described below.

[1124] System Configuration

[1125] server

[1126] The server includes the following functions:

[1127] 1. Data collection method: Use tools to collect information such as the manager's past statements, books, and video recordings. For this purpose, web scraping and APIs can be used. Voice recognition technology is applied to the video data to convert it into text data.

[1128] 2. Data cleaning method: The collected data is cleaned, noise is removed, and the format is standardized. The text data is tokenized and segmented into sentences, and then stored in a database.

[1129] 3. Generative AI model training method: Train a generative AI model (e.g., GPT-2) using the preprocessed data. Train the model using a deep learning framework (e.g., TensorFlow) and improve its performance through training and testing.

[1130] 4. Advice provision method: Generative AI models are used to generate appropriate advice for users' management inquiries. Scenario-based risk assessments and strategy proposals are provided in real time.

[1131] 5. Model Update Method: Regularly update the generative AI model with new data and feedback. Keep the model up to date by adding newly acquired data (e.g., interview or lecture data).

[1132] Terminal

[1133] The terminal includes the following features:

[1134] 1. Interface provision means: Provide an interface for virtual store operators to receive management decision advice in real time through smart glasses.

[1135] 2. Result display means: Displays the results of management decisions and advice. Depending on the scenario entered by the user, specific risk assessments and strategy proposals are displayed in text format.

[1136] User

[1137] The user does the following:

[1138] 1. Scenario input: Enter a specific business situation or scenario into the terminal. For example, enter a prompt such as, "Please tell us your risk assessment regarding the launch of a new product into the market."

[1139] 2. Refer to the results: Refer to the advice and evaluations provided by the generative AI model returned from the server and make actual management decisions.

[1140] Specific examples

[1141] Consider a case where a virtual store operator uses smart glasses to have a risk assessment performed by a generative AI model when introducing a new product. When the user types "Please tell me the risk assessment for the market launch of a new product" into the device interface, the server uses the generative AI model to generate an appropriate risk assessment and displays the result in real time on the smart glasses. This allows the operator to make an appropriate decision quickly on the spot.

[1142] As described above, the present invention provides virtual store operators with appropriate advice on business decisions in real time, thereby realizing more efficient operations.

[1143] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1144] Step 1:

[1145] The server collects data such as the manager's past statements, books, and video recordings. The input is data obtained from web scraping tools and APIs, and the output is text data or raw data before cleaning. This data is converted into text format using voice recognition technology (e.g., Google Speech-to-Text) or OCR.

[1146] Step 2:

[1147] The server cleans the collected data and standardizes its format. The input is the text data or raw data obtained in step 1, and the output is text data that has been denoised and integrated into a unified format. Specific operations include deleting duplicate data, removing unnecessary noise, tokenising (word splitting) and splitting sentences.

[1148] Step 3:

[1149] The server trains a generative AI model (e.g., GPT-2) using the preprocessed data. The input is the preprocessed text data, and the output is the trained generative AI model. Specifically, a deep learning framework (e.g., TensorFlow) is used to divide the text data into a training set and a test set, and to train the model and evaluate its performance.

[1150] Step 4:

[1151] The user inputs a specific business situation or scenario into the terminal interface. The input is a prompt, for example, "Please tell me your risk assessment regarding the launch of a new product into the market." The input prompt is then sent to the server.

[1152] Step 5:

[1153] The server uses a generative AI model to generate business decisions and advice based on the prompt text entered by the user. The input is the data from Step 4, which includes the prompt text and the generative AI model, and the output is the generated business decisions and advice text. The model generates optimal risk assessments and strategic proposals based on the input prompt text.

[1154] Step 6:

[1155] The terminal displays the management decisions and advice received from the server to the user. The input is the generated results sent from the server, and the output is text information displayed on a device such as smart glasses. Specifically, the user can check the advice in real time through the display of the smart glasses.

[1156] Step 7:

[1157] The server periodically updates the generative AI model using new data and user feedback. The input is the newly collected data or feedback, and the output is the updated generative AI model. This ensures that the model's accuracy and performance are always kept up to date.

[1158] 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.

[1159] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[1160] Data collection

[1161] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[1162] Data Preprocessing

[1163] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[1164] Model training

[1165] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[1166] Business decision simulation

[1167] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[1168] Management Consulting Platform

[1169] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[1170] Introducing the Emotion Engine

[1171] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module that analyzes voice data and an image analysis module that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[1172] Model emotional response

[1173] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[1174] Model Update

[1175] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[1176] Specific examples

[1177] Examples of data collection

[1178] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[1179] Specific examples of business decision simulation

[1180] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[1181] Examples of management consulting platforms

[1182] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[1183] Examples of emotion engines

[1184] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[1185] The above is an embodiment of the present invention. By using this system, efficient and appropriate business decisions and business consultations can be made while inheriting the knowledge of the business owner, and more personalized advice can be provided that responds to the user's emotions.

[1186] The processing flow will be explained below.

[1187] Data collection

[1188] Step 1:

[1189] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[1190] Step 2:

[1191] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[1192] Step 3:

[1193] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[1194] Data Preprocessing

[1195] Step 1:

[1196] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[1197] Step 2:

[1198] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[1199] Step 3:

[1200] The server converts the cleaned data into a unified format and stores it in a database.

[1201] Model training

[1202] Step 1:

[1203] The server splits the preprocessed dataset into a training set and a test set.

[1204] Step 2:

[1205] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[1206] Step 3:

[1207] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[1208] Step 4:

[1209] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[1210] Business decision simulation

[1211] Step 1:

[1212] The terminal provides an interface to the simulation tool for internal users.

[1213] Step 2:

[1214] The user inputs a specific business scenario (e.g., new market entry, product development).

[1215] Step 3:

[1216] The server receives the input scenario and queries the generative AI model.

[1217] Step 4:

[1218] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[1219] Step 5:

[1220] The server returns the results to the terminal, where the user views the results.

[1221] Management Consulting Platform

[1222] Step 1:

[1223] The terminal provides an interface for the management consulting platform to external users.

[1224] Step 2:

[1225] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[1226] Step 3:

[1227] The server receives the input consultation content and executes a query against the generative AI model.

[1228] Step 4:

[1229] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[1230] Step 5:

[1231] The server sends advice and solutions to the terminal, which the user then confirms.

[1232] Introducing the Emotion Engine

[1233] Step 1:

[1234] The terminal activates a microphone for capturing the user's voice data and a camera for capturing facial expression data.

[1235] Step 2:

[1236] The server uses a voice analysis module to analyze the captured voice data and recognize the user's emotions (e.g., joy, sadness, anger, surprise).

[1237] Step 3:

[1238] The server uses an image analysis module to analyze the captured facial expression data and recognize the user's emotions.

[1239] Model emotional response

[1240] Step 1:

[1241] The server receives the user's emotional data analyzed by the emotion engine and determines the user's current emotional state.

[1242] Step 2:

[1243] The server adjusts the output of the generative AI model based on the determined emotion data. For example, if the user is nervous, the server adjusts the advice content to be gentler.

[1244] Step 3:

[1245] The generative AI model generates emotion-based tailored advice and business decisions and sends the results to a server.

[1246] Step 4:

[1247] The server sends the adjusted results to the terminal, and the user confirms them.

[1248] Model Update

[1249] Step 1:

[1250] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[1251] Step 2:

[1252] The server cleans the new data and standardizes the format.

[1253] Step 3:

[1254] The server uses the updated data to retrain the generative AI model and emotion engine.

[1255] Step 4:

[1256] The server evaluates the performance of the retrained model and releases an update.

[1257] The above are the specific processing steps of the program for the system that combines the emotion engine.

[1258] Example 2

[1259] 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."

[1260] Conventional management support systems have difficulty in fully utilizing the manager's past knowledge and experience, and in providing advice that reflects the user's feelings. As a result, they are unable to provide efficient and appropriate support for management decisions and management consultations, resulting in problems such as lower user satisfaction and success rates.

[1261] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for collecting the manager's past words and actions, books, video recordings, etc., a means for cleaning the collected data and standardizing the format, and a means for training the generative AI model using the preprocessed data. This makes it possible to provide efficient and appropriate management decisions and management consultations while inheriting the manager's knowledge.

[1262] "Past words and actions of management" refers to past words and actions made by management, as well as records of those actions.

[1263] "Books" refers to books written by business managers and books on management.

[1264] "Video recordings" refers to video or recorded media such as speeches or interviews in which the executive appears.

[1265] "Collection methods" refers to the technologies and methods used to obtain relevant data from the internet or physical media.

[1266] "Cleaning and formatting methods" refers to techniques and methods used to remove unnecessary information from collected data and convert it into a format suitable for analysis.

[1267] A "generative AI model" refers to an artificial intelligence model that is trained to reproduce the actions and behavior of managers.

[1268] "Preprocessed data" refers to data that has been cleaned and formatted and is used to train a model.

[1269] "Training means" refers to techniques and methods, including learning algorithms and computing resources, for improving the accuracy of artificial intelligence models.

[1270] "Business situations and scenarios" refer to specific business decision-making situations and strategies that users are considering.

[1271] "Means for generating business decisions" refers to technologies and methods for deriving optimal business decisions based on scenarios input by users using generative AI models.

[1272] "Advice using a generative AI model for business consultations" refers to advice and suggestions generated by AI in response to business-related questions and inquiries from users.

[1273] "Means of updating generative AI models with new data and feedback" refers to techniques and methods that incorporate new data collected and user feedback to improve the accuracy of the model and keep it up to date with the latest information.

[1274] An "emotion-recognizing emotion engine" refers to an artificial intelligence module that analyzes the user's emotional state from their voice and facial expressions.

[1275] "Means for adjusting output" refers to techniques or methods for changing the content or tone of the output of a generative AI model based on the user's emotional state.

[1276] "Speech recognition technology" refers to technology for converting voice data into text data.

[1277] "Optical Character Recognition (OCR)" refers to technology for extracting character data from images and scanned documents.

[1278] "Database" refers to a digital storage system that stores collected data in an organized manner and makes it easily searchable and accessible.

[1279] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[1280] Data collection

[1281] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the Internet. It then automatically converts video data into text using voice recognition technology (e.g., Google Speech-to-Text API), and digitizes book and article data using optical character recognition (OCR) technology (e.g., Tesseract OCR).

[1282] Data Preprocessing

[1283] The server then cleans and standardizes the collected data by removing duplicates and noise, tokenizing the text data (splitting words and sentences), and storing it in a consistent format in a database (e.g., MySQL, PostgreSQL).

[1284] Model training

[1285] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[1286] Business decision simulation

[1287] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[1288] Management Consulting Platform

[1289] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[1290] Introducing the Emotion Engine

[1291] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module (e.g., OpenSMILE) that analyzes voice data and an image analysis module (e.g., OpenCV, Dlib) that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[1292] Model emotional response

[1293] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[1294] Model Update

[1295] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[1296] Specific examples

[1297] Examples of data collection

[1298] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[1299] Specific examples of business decision simulation

[1300] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[1301] Examples of management consulting platforms

[1302] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[1303] Examples of emotion engines

[1304] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[1305] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management, and it is possible to provide more personalized advice that responds to the user's emotions.

[1306] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1307] Program processing flow

[1308] Step 1: Data collection

[1309] The server collects the manager's past statements, actions, books, and video recordings from the internet. Specifically, it obtains the data using web scraping tools and APIs. Video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and book and article data is digitized using optical character recognition (OCR) technology (e.g., Tesseract OCR). The input is URLs and file paths related to the manager, and the output is a collection of text data.

[1310] Step 2: Data Preprocessing

[1311] The server cleans the collected data and standardizes its format. Specifically, it removes duplicate data and noise. It tokenizes (splits words) and sentences the text data and stores it in a consistent format in a database (e.g., MySQL, PostgreSQL). The input is the text data collected in step 1, and the output is organized data in a unified format.

[1312] Step 3: Model training

[1313] The server uses the preprocessed data to train a generative AI model. A deep learning framework (e.g., TensorFlow or PyTorch) is used to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning. The input is the preprocessed data from step 2, and the output is a trained generative AI model.

[1314] Step 4: Business decision simulation

[1315] The terminal provides an interface that allows internal users to input specific business situations and scenarios. Users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives the scenarios and simulates business decisions using a generative AI model. The results are sent back to the terminal, and the user refers to them to make the final business decision. The input is the business scenario, and the output is the simulation results.

[1316] Step 5: Management Consulting Platform

[1317] The terminal provides an interface for external users to receive management consultations. Users input specific questions about "fundraising strategies" and "recruitment methods." The server receives the input consultation content, generates appropriate advice using a generative AI model, and displays the output on the terminal. The user uses this advice as a reference when making management decisions. The input is the consultation content, and the output is the generated advice.

[1318] Step 6: Implementing the Emotion Engine

[1319] The device is equipped with a camera and microphone to capture the user's voice and facial expressions in real time. The server analyzes the voice data using a voice analysis module (e.g., OpenSMILE) and the facial expression data using an image analysis module (e.g., OpenCV, Dlib). The input is voice and facial expression data, and the output is the emotional data resulting from the analysis.

[1320] Step 7: Model emotional response

[1321] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, it will provide gentler advice, and if positive emotions are recognized, it will provide more proactive advice. The input is emotional data, and the output is the adjusted advice or judgment.

[1322] Step 8: Model Update

[1323] The server periodically updates the generative AI model based on new data and feedback. It adds newly collected interview and lecture data, integrates it with the existing dataset, and retrains the model. The input is the new data and feedback, and the output is an updated generative AI model.

[1324] The above are the specific processing steps of the system program. The operations performed at each step enable the user to make more efficient and appropriate business decisions and receive advice.

[1325] (Application example 2)

[1326] 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."

[1327] In today's business environment, it is difficult to pass on the knowledge and experience of managers to future generations, and there is a need for real-time management decisions and consultations that respond to users' emotions. Monitoring employee mental health is also an important issue, and it is difficult to provide effective countermeasures at the appropriate time. For this reason, there is a need to develop a system that can reproduce the words and actions of managers and provide real-time advice that responds to users' emotions.

[1328] 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 the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using preprocessed data, means for generating management decisions based on specific management situations and scenarios, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, an emotion recognition engine for recognizing the user's emotions, means for adjusting the output of the generative AI model based on emotion data using the emotion recognition engine, means for analyzing real-time data acquired from sensors, and means for providing appropriate responses and advice in real time using the generative AI model. This enables the provision of personalized advice that corresponds to the user's emotions and real-time mental health monitoring of employees while inheriting the manager's knowledge.

[1329] "Past words and actions of management" refers to records of statements, decisions, actions, etc. made by management in the past.

[1330] "Books" are books written by business managers or management reference works.

[1331] "Video recordings" refers to video data such as videos, interviews, and lectures featuring management.

[1332] "Means of collecting data" refers to methods of obtaining data, such as using web scraping tools, APIs, OCR technology, etc.

[1333] "Means for cleaning data and standardizing the format" refers to methods for removing duplication and noise from collected data, and for tokenizing and segmenting text data.

[1334] "Means for training a generative AI model using preprocessed data" refers to a method for training a generative AI model using cleaned data using a deep learning framework (e.g., TensorFlow or PyTorch).

[1335] A "means for generating business decisions" is a method for using a generative AI model to form business decisions based on specific business situations or scenarios.

[1336] The "means of providing advice" is a method of providing advice to a user's business consultation using a generative AI model.

[1337] A "generative AI model" is a model that uses deep learning technology to reproduce the words, actions, and thoughts of managers and generate management decisions and advice.

[1338] "Feedback-based updating" is a method of retraining a model based on new data and user feedback to improve its performance.

[1339] An "emotion recognition engine" is an engine that uses a voice analysis module and an image analysis module to recognize a user's emotions in real time.

[1340] "Means for adjusting output based on emotional data" refers to a method for adjusting the output results of a generative AI model based on emotional data analyzed by an emotion recognition engine.

[1341] "Means for analyzing real-time data" refers to methods for analyzing data obtained from security cameras and audio sensors in real time.

[1342] "Means for providing appropriate responses and advice in real time" refers to methods for providing appropriate responses and advice on the spot based on data analyzed using a generative AI model.

[1343] The present invention relates to a system that combines an emotion recognition engine and a generative AI model to provide management decisions based on the manager's past words and actions and experience, as well as real-time advice corresponding to the user's emotions. This system aims to make decisions in specific management situations and manage the mental health of employees. A specific embodiment of the present invention is described below.

[1344] Hardware and software used

[1345] 1. Hardware:

[1346] security cameras

[1347] Audio Sensor

[1348] microphone

[1349] GPU-equipped servers

[1350] 2. Software:

[1351] Python

[1352] OpenCV (image analysis library)

[1353] Keras (deep learning framework)

[1354] TensorFlow (deep learning framework)

[1355] SpeechRecognition library (audio analysis)

[1356] GPT-2 model (generative AI model)

[1357] Specific Embodiments of the System

[1358] 1. Data Collection

[1359] The server collects information such as past statements and actions of executives, books, and video recordings from online platforms and internal databases. The collected video data is converted into text using voice recognition technology, and books and articles are converted into digital data using OCR technology.

[1360] 2. Data Preprocessing

[1361] The server cleans the collected data, removes duplicate data and noise, tokenizes the text data, splits it into sentences, and stores it in a unified database.

[1362] 3. Model training

[1363] The server uses the preprocessed data to train a generative AI model. A deep learning framework (TensorFlow or PyTorch) is used to create the generative AI model. The model is divided into a training set and a test set, and its performance is improved through reinforcement learning and supervised learning.

[1364] 4. Emotion recognition

[1365] It analyzes real-time data (audio and video) of employees and users acquired through security cameras and audio sensors, and detects user emotions using an emotion recognition engine. It uses the SpeechRecognition library for audio analysis and OpenCV for image analysis.

[1366] 5. Tuning generative AI models based on emotion data

[1367] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion recognition engine. For example, if the user is feeling stressed, the generative AI model will be adjusted to output gentle and calming advice.

[1368] 6. Real-time advice

[1369] The server generates appropriate responses and advice based on real-time data analyzed using the generative AI model and displays them on the user's device, making it possible to monitor employee mental health and support management decision-making.

[1370] Specific examples

[1371] The server downloads videos of past executive speeches from an online platform, converts them into text using speech recognition technology, and stores them in a database after a cleaning process. An in-house project manager accesses a simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, which are then sent back to the terminal. A user using the terminal asks a question by voice, and the server uses a speech analysis module to analyze the user's emotions and determine that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[1372] Prompt Sentence Examples

[1373] "Generate examples of measures to take when an employee is under stress."

[1374] A system based on this format will utilize the knowledge of management, provide advice that responds to the user's emotions, and enable real-time mental health monitoring of employees.

[1375] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1376] Step 1:

[1377] The server collects information about the manager's past statements, actions, books, video recordings, etc. from online platforms and internal databases. Specifically, it acquires data using web scraping tools, APIs, and OCR technology. The inputs are the URL of the online platform and a database query, and the acquired data (text, audio, video) is output.

[1378] Step 2:

[1379] The server cleans the collected data by removing duplicates, removing noise, tokenizing text data, and segmenting sentences. The input is the collected raw data, and the output is cleaned, consistent data.

[1380] Step 3:

[1381] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework (TensorFlow or PyTorch) and performs reinforcement learning or supervised learning using a training set and a test set. It takes the preprocessed data as input and outputs a trained generative AI model.

[1382] Step 4:

[1383] A user accesses the business decision simulation tool using a terminal and inputs a specific business situation or scenario, such as "entering a new market" or "cost reduction strategy." The scenario information is input, and a simulation request is sent to the server.

[1384] Step 5:

[1385] The server uses a generative AI model to generate business decisions based on the specified scenario. Specifically, it analyzes the input scenario information and generates business decisions and strategic proposals. A scenario request is input, and the generated business decisions are output and sent to the terminal.

[1386] Step 6:

[1387] A user using a terminal accesses the business consulting platform and inputs a specific consultation content (e.g., "Fundraising strategy for business expansion"). With the consultation content as input, a consultation request is sent to the server.

[1388] Step 7:

[1389] The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. A consultation request is input, and the generated advice is output.

[1390] Step 8:

[1391] The device uses security cameras and audio sensors to capture real-time audio and video data to analyze the user's emotions. An emotion recognition engine is used to analyze the audio and facial expression data to identify the user's emotional state. Real-time data is input, and the emotion analysis results are output.

[1392] Step 9:

[1393] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion recognition engine. Specifically, if the user is nervous, it provides gentle and calm advice, and if positive emotions are recognized, it provides proactive advice. The input is the emotion analysis result, and the adjusted advice is output.

[1394] Step 10:

[1395] The server uses the generative AI model to provide appropriate responses and advice in real time, which are displayed on the device. This enables monitoring of employee mental health and supporting management decisions. The input is adjusted advice, and the final output is real-time responses and advice displayed on the user's device.

[1396] This realizes an efficient system that responds to the specific prompt statement mentioned above, "Please generate examples of measures to be taken when an employee is under stress."

[1397] 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.

[1398] 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.

[1399] 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.

[1400] [Fourth embodiment]

[1401] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1402] 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.

[1403] 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).

[1404] 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.

[1405] 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.

[1406] 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).

[1407] 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.

[1408] 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.

[1409] 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.

[1410] 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.

[1411] 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.

[1412] 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.

[1413] 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."

[1414] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[1415] Data collection

[1416] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[1417] Data Preprocessing

[1418] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[1419] Model training

[1420] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[1421] Business decision simulation

[1422] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[1423] Management Consulting Platform

[1424] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[1425] Model Update

[1426] The server periodically updates the generative AI model using new data and feedback, for example by collecting newly acquired executive interviews or speeches and adding them to the existing dataset, ensuring that the generative AI model always provides decisions based on the most up-to-date information.

[1427] Specific examples

[1428] Examples of data collection

[1429] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[1430] Specific examples of business decision simulation

[1431] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[1432] Examples of management consulting platforms

[1433] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[1434] The above is an embodiment of the present invention. By using this system, it becomes possible to inherit the knowledge of the manager and to make efficient and appropriate management decisions and provide management consultations.

[1435] The processing flow will be explained below.

[1436] Data collection

[1437] Step 1:

[1438] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[1439] Step 2:

[1440] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[1441] Step 3:

[1442] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[1443] Data Preprocessing

[1444] Step 1:

[1445] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[1446] Step 2:

[1447] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[1448] Step 3:

[1449] The server converts the cleaned data into a unified format and stores it in a database.

[1450] Model training

[1451] Step 1:

[1452] The server splits the preprocessed dataset into a training set and a test set.

[1453] Step 2:

[1454] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[1455] Step 3:

[1456] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[1457] Step 4:

[1458] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[1459] Business decision simulation

[1460] Step 1:

[1461] The terminal provides an interface to the simulation tool for internal users.

[1462] Step 2:

[1463] The user inputs a specific business scenario (e.g., new market entry, product development).

[1464] Step 3:

[1465] The server receives the input scenario and queries the generative AI model.

[1466] Step 4:

[1467] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[1468] Step 5:

[1469] The server returns the results to the terminal, where the user views the results.

[1470] Management Consulting Platform

[1471] Step 1:

[1472] The terminal provides an interface for the management consulting platform to external users.

[1473] Step 2:

[1474] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[1475] Step 3:

[1476] The server receives the input consultation content and executes a query against the generative AI model.

[1477] Step 4:

[1478] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[1479] Step 5:

[1480] The server sends advice and solutions to the terminal, which the user then confirms.

[1481] Model Update

[1482] Step 1:

[1483] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[1484] Step 2:

[1485] The server cleans the new data and standardizes the format.

[1486] Step 3:

[1487] The server retrains the generative AI model with the updated data.

[1488] Step 4:

[1489] The server evaluates the performance of the retrained model and releases an update.

[1490] The above are the specific steps of the system program processing.

[1491] Example 1

[1492] 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."

[1493] There is a need for a system that can automatically and efficiently provide appropriate advice by systematizing management decisions that reflect the past words, actions, and thoughts of managers. However, current systems require cumbersome data collection and preprocessing, making it difficult to train effective generative AI models. Furthermore, users lack a means to instantly obtain appropriate management decisions and advice for specific scenarios. Furthermore, the process of updating the model based on new data and feedback is not automated, making it difficult to keep up with the latest management information.

[1494] 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.

[1495] In this invention, the server includes means for collecting the manager's past words and actions, literature, video recordings, etc., means for cleaning the collected data and standardizing the format, and means for training the generative AI model using the preprocessed data. This enables means for a user to input a specific management scenario and generate management decisions based on that scenario, means for providing advice using the generative AI model in response to a user's management consultation, and means for updating the generative AI model using new data and feedback.

[1496] A "management officer" is a person who is responsible for making important decisions in the operation of a company or organization.

[1497] "Behavior" refers collectively to speech and actions, and refers to words and actions shown in specific situations or on specific themes.

[1498] "Literature" is any kind of written material written to provide information, such as books, articles, or reports.

[1499] A "visual recording" is a collection of visual and audio information stored in the form of video, video recording or other media.

[1500] "Data collection" is the process of obtaining and gathering information needed for a specific purpose.

[1501] "Cleaning" is the process of removing errors and noise from collected data, organizing and normalizing the data.

[1502] "Unifying formats" is the process of converting data with different formats and structures into a consistent, standard format.

[1503] "Preprocessing" refers to data preparation performed before data analysis or model training.

[1504] A "generative AI model" is an algorithm that uses artificial intelligence techniques to learn specific patterns and characteristics and make inferences based on new data.

[1505] "Training" is the process of teaching a generative AI model using data to improve its performance.

[1506] A "business scenario" is a setting that represents assumptions and conditions regarding a particular business situation or strategy.

[1507] "Management decision-making" is the process by which managers make rational decisions regarding the operation of a company or organization.

[1508] "Management consultation" is the act of seeking professional advice on management issues and problems.

[1509] "Advice" is helpful advice or direction given on a specific problem or issue.

[1510] "Feedback" is evaluation or information provided to improve the performance of a system or model.

[1511] "Updating" is the process of adding new information or data to an existing system or model to improve its performance or accuracy.

[1512] A "database" is an organized collection of data that can be efficiently managed and searched.

[1513] A "user terminal" is a device or interface that a user can directly access and operate.

[1514] overview

[1515] The present invention relates to a system for making business decisions and providing business advice using a generative AI model that has learned the words, actions, and thoughts of business managers. Specific embodiments of this system are described in detail below.

[1516] Data collection

[1517] The server first collects information on the manager's past statements and actions, documents, video recordings, etc. Tools used include web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). Furthermore, the video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and the document data is digitized using OCR technology (e.g., Tesseract).

[1518] Data Preprocessing

[1519] The server cleans the collected data and standardizes its format. The cleaning process involves removing duplicate data and removing noise. Tools used include NLP libraries (e.g., NLTK, spaCy). This data processing involves tokenizing the text data (splitting it into words) and storing it in a consistent format in the database.

[1520] Model training

[1521] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework such as TensorFlow or PyTorch. The dataset is divided into a training set and a test set, and the model's performance is evaluated and improved through cross-validation.

[1522] Business decision simulation

[1523] The terminal provides an interface that users can access and input specific business scenarios. For example, a web application may be provided that provides a form where users can input scenarios such as "entering a new market."

[1524] Users input scenarios through a simulation tool, and the server inputs prompt statements (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate business decisions.

[1525] The server sends the generated business decisions back to the terminal, allowing the user to view the results, for example by displaying the results on a dashboard so the user can view the details.

[1526] Management Consulting Platform

[1527] The terminal provides an interface for external users to receive management consultations. For example, it provides a chatbot and a consultation form, allowing users to input their consultation details, such as "fundraising strategies" or "recruitment methods."

[1528] The user inputs a specific business consultation, and the server generates advice using a generative AI model.

[1529] The server sends the generated advice back to the terminal, and the user makes management decisions based on that information. The advice is displayed in a chat window or in a report format.

[1530] Model Update

[1531] The server periodically updates the generative AI model with new data and feedback, for example by adding newly acquired executive interviews or speech data and updating the existing dataset.

[1532] The server retrains the model with the latest data to optimize its performance.

[1533] Specific examples

[1534] Examples of data collection

[1535] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[1536] Specific examples of business decision simulation

[1537] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[1538] Examples of management consulting platforms

[1539] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[1540] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management.

[1541] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1542] Step 1: Data collection

[1543] The server collects data related to the manager. Input includes the manager's lecture videos, blog posts, interview articles, etc. from online platforms. Specific operations include using web scraping tools (e.g., BeautifulSoup, Scrapy) and APIs (e.g., YouTube API). These tools are used to automatically collect the necessary data and save it as a file. The output is raw data stored in local storage.

[1544] Step 2: Data Preprocessing

[1545] The server preprocesses the collected data. The input includes the raw data collected in step 1. Specifically, it converts voice data into text using speech recognition technology (e.g., Google Speech-to-Text API), and converts image and PDF data into digital text using OCR technology (e.g., Tesseract). It also performs noise removal, de-duplicate data removal, and tokenization using NLP libraries (e.g., NLTK, spaCy). The output is text data in a clean, unified format.

[1546] Step 3: Data storage

[1547] The server stores the preprocessed data in a database. The input includes the clean text data generated in step 2. Specifically, it uses a database management system (e.g., MySQL, PostgreSQL) to store the text data in a structured format in the database. The output is the preprocessed text data stored in the database.

[1548] Step 4: Model training

[1549] The server uses the preprocessed data to train a generative AI model. The input includes the text data stored in the database in step 3. Specifically, a model is built using a deep learning framework (e.g., TensorFlow, PyTorch) and supervised learning and reinforcement learning are performed. The dataset is divided into a training set and a test set, and the model's performance is evaluated and optimized through cross-validation. The output is a generative AI model that can reproduce the words, actions, and thoughts of the manager.

[1550] Step 5: Business decision simulation

[1551] The terminal provides an interface for the user to input a business scenario. The input includes the business scenario (e.g., "Enter a new market") entered by the user. Specifically, it provides a web form or dashboard to allow the user to input the scenario. The server receives this scenario and inputs a prompt statement (e.g., "What risks should be considered when deciding to enter a new market?") into the generative AI model to generate a business decision. The output is the generated business decision, which is sent back to the terminal and displayed to the user.

[1552] Step 6: Management Consulting Platform

[1553] The terminal provides an interface for external users to input business advice. The input includes the specific advice entered by the user (e.g., "fundraising strategy"). Specific operations include providing a chatbot and a consultation form, allowing the user to input the details of the consultation. The server receives the consultation details and generates appropriate advice using a generative AI model. The output is the generated advice, which is sent back to the terminal and displayed to the user.

[1554] Step 7: Model Update

[1555] The server periodically updates the generative AI model using new data and feedback. The input includes newly collected managerial utterance data and feedback data. Specifically, it adds new data to the old dataset and retrains the model to optimize its performance. The output is an updated generative AI model, which can provide management decisions and advice based on the latest information.

[1556] (Application example 1)

[1557] 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."

[1558] In modern virtual store operations, managers must make a wide range of management decisions quickly and appropriately. However, relying on the manager's personal knowledge and experience can lead to biased decisions, resulting in insufficient risk assessment and strategic proposals. Furthermore, traditional management support systems have difficulty providing advice in real time, making them less effective in situations where immediate decisions are required on-site. This increases the burden on virtual store operators and hinders efficient operations, creating the challenge.

[1559] 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.

[1560] In this invention, the server includes means for collecting the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using the preprocessed data, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, means for providing management decision advice to the virtual store manager in real time via smart glasses, and means for generating risk assessments and strategic proposals based on scenarios specified by the user. This enables the virtual store manager to receive advice from the generative AI model in real time and make quick and appropriate management decisions.

[1561] A "management officer" is a person responsible for making decisions regarding the operation of a company or organization.

[1562] "Behavior" refers to all behavior, including verbal expressions and actions.

[1563] A "book" is a collection of bound documents that contain printed text and / or images and have a definite form.

[1564] "Visual recording" refers to data containing visual information in the form of video, film, etc.

[1565] "Collection methods" refer to methods and techniques for systematically gathering specific information or data.

[1566] "Cleaning" refers to the process of removing noise from data and arranging it into a unified format.

[1567] A "generative AI model" is an artificial intelligence model that learns using machine learning algorithms based on collected data.

[1568] "Training methods" refer to methods or techniques for training a generative AI model using a specific dataset.

[1569] "Business condition" refers to the current performance and environmental state of a company or organization.

[1570] A "scenario" refers to a hypothetical series of events or happenings based on specific circumstances or conditions.

[1571] "Advice delivery means" refers to the methods and technologies for communicating business decisions and proposals obtained by generative AI models to users.

[1572] "Update methods" refer to methods and techniques that incorporate new data and feedback into generative AI models to improve their accuracy and performance.

[1573] A "virtual store" refers to a store that is not dependent on a physical location and is operated on the Internet.

[1574] "Smart glasses" refers to a glasses-type device equipped with a display and sensors to provide visual information to the wearer.

[1575] "Real-time" refers to a state in which processing or response is carried out immediately the moment a specific operation or event occurs.

[1576] "Risk assessment" refers to the process of analyzing and evaluating the potential hazards and uncertainties associated with a particular action or scenario.

[1577] A "strategic proposal" refers to the presentation of a plan or method designed to achieve a specific objective.

[1578] "User" refers to a person such as a manager or entrepreneur who uses this system.

[1579] "Database" refers to a collection of data organized in a particular way and designed to be efficiently managed and searched.

[1580] "Terminal" refers to a device that is connected to a system or network and allows a user to input and output information.

[1581] The present invention relates to a system for providing support to a manager in making quick and appropriate management decisions in the operation of a virtual store. Specific embodiments for carrying out the present invention are described below.

[1582] System Configuration

[1583] server

[1584] The server includes the following functions:

[1585] 1. Data collection method: Use tools to collect information such as the manager's past statements, books, and video recordings. For this purpose, web scraping and APIs can be used. Voice recognition technology is applied to the video data to convert it into text data.

[1586] 2. Data cleaning method: The collected data is cleaned, noise is removed, and the format is standardized. The text data is tokenized and segmented into sentences, and then stored in a database.

[1587] 3. Generative AI model training method: Train a generative AI model (e.g., GPT-2) using the preprocessed data. Train the model using a deep learning framework (e.g., TensorFlow) and improve its performance through training and testing.

[1588] 4. Advice provision method: Generative AI models are used to generate appropriate advice for users' management inquiries. Scenario-based risk assessments and strategy proposals are provided in real time.

[1589] 5. Model Update Method: Regularly update the generative AI model with new data and feedback. Keep the model up to date by adding newly acquired data (e.g., interview or lecture data).

[1590] Terminal

[1591] The terminal includes the following features:

[1592] 1. Interface provision means: Provide an interface for virtual store operators to receive management decision advice in real time through smart glasses.

[1593] 2. Result display means: Displays the results of management decisions and advice. Depending on the scenario entered by the user, specific risk assessments and strategy proposals are displayed in text format.

[1594] User

[1595] The user does the following:

[1596] 1. Scenario input: Enter a specific business situation or scenario into the terminal. For example, enter a prompt such as, "Please tell us your risk assessment regarding the launch of a new product into the market."

[1597] 2. Refer to the results: Refer to the advice and evaluations provided by the generative AI model returned from the server and make actual management decisions.

[1598] Specific examples

[1599] Consider a case where a virtual store operator uses smart glasses to have a risk assessment performed by a generative AI model when introducing a new product. When the user types "Please tell me the risk assessment for the market launch of a new product" into the device interface, the server uses the generative AI model to generate an appropriate risk assessment and displays the result in real time on the smart glasses. This allows the operator to make an appropriate decision quickly on the spot.

[1600] As described above, the present invention provides virtual store operators with appropriate advice on business decisions in real time, thereby realizing more efficient operations.

[1601] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1602] Step 1:

[1603] The server collects data such as the manager's past statements, books, and video recordings. The input is data obtained from web scraping tools and APIs, and the output is text data or raw data before cleaning. This data is converted into text format using voice recognition technology (e.g., Google Speech-to-Text) or OCR.

[1604] Step 2:

[1605] The server cleans the collected data and standardizes its format. The input is the text data or raw data obtained in step 1, and the output is text data that has been denoised and integrated into a unified format. Specific operations include deleting duplicate data, removing unnecessary noise, tokenising (word splitting) and splitting sentences.

[1606] Step 3:

[1607] The server trains a generative AI model (e.g., GPT-2) using the preprocessed data. The input is the preprocessed text data, and the output is the trained generative AI model. Specifically, a deep learning framework (e.g., TensorFlow) is used to divide the text data into a training set and a test set, and to train the model and evaluate its performance.

[1608] Step 4:

[1609] The user inputs a specific business situation or scenario into the terminal interface. The input is a prompt, for example, "Please tell me your risk assessment regarding the launch of a new product into the market." The input prompt is then sent to the server.

[1610] Step 5:

[1611] The server uses a generative AI model to generate business decisions and advice based on the prompt text entered by the user. The input is the data from Step 4, which includes the prompt text and the generative AI model, and the output is the generated business decisions and advice text. The model generates optimal risk assessments and strategic proposals based on the input prompt text.

[1612] Step 6:

[1613] The terminal displays the management decisions and advice received from the server to the user. The input is the generated results sent from the server, and the output is text information displayed on a device such as smart glasses. Specifically, the user can check the advice in real time through the display of the smart glasses.

[1614] Step 7:

[1615] The server periodically updates the generative AI model using new data and user feedback. The input is the newly collected data or feedback, and the output is the updated generative AI model. This ensures that the model's accuracy and performance are always kept up to date.

[1616] 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.

[1617] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[1618] Data collection

[1619] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the internet. It then automatically converts video data into text using voice recognition technology, and digitizes book and article data using OCR (optical character recognition) technology.

[1620] Data Preprocessing

[1621] The server then cleans the collected data and standardizes its format, specifically removing duplicate data and removing noise from the data, tokenizing the text data (segmenting words and sentences), and storing it in a consistent format in the database.

[1622] Model training

[1623] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[1624] Business decision simulation

[1625] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[1626] Management Consulting Platform

[1627] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[1628] Introducing the Emotion Engine

[1629] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module that analyzes voice data and an image analysis module that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[1630] Model emotional response

[1631] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[1632] Model Update

[1633] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[1634] Specific examples

[1635] Examples of data collection

[1636] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[1637] Specific examples of business decision simulation

[1638] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[1639] Examples of management consulting platforms

[1640] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[1641] Examples of emotion engines

[1642] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[1643] The above is an embodiment of the present invention. By using this system, efficient and appropriate business decisions and business consultations can be made while inheriting the knowledge of the business owner, and more personalized advice can be provided that responds to the user's emotions.

[1644] The processing flow will be explained below.

[1645] Data collection

[1646] Step 1:

[1647] The server retrieves text data about business managers from the Internet using web scraping tools and APIs.

[1648] Step 2:

[1649] The server downloads the executive's speech video from an online platform, then uses a speech recognition API (e.g., Google Speech-to-Text) to extract the audio from the video and convert it into text data.

[1650] Step 3:

[1651] The server converts text data such as books and newspaper articles into digital data using OCR (optical character recognition) technology.

[1652] Data Preprocessing

[1653] Step 1:

[1654] The server removes duplicate content from the collected text data by normalizing the text and matching it with existing data in the database.

[1655] Step 2:

[1656] The server tokenizes (splits words) and sentences into text data, and also cleans it by removing unnecessary metadata and noise.

[1657] Step 3:

[1658] The server converts the cleaned data into a unified format and stores it in a database.

[1659] Model training

[1660] Step 1:

[1661] The server splits the preprocessed dataset into a training set and a test set.

[1662] Step 2:

[1663] The server uses a deep learning framework (e.g., TensorFlow or PyTorch) to design the network structure of the generative AI model.

[1664] Step 3:

[1665] The server uses reinforcement learning and supervised learning to train the model, allowing it to learn the manager's thinking and decision-making criteria.

[1666] Step 4:

[1667] The server evaluates the performance of the model and adjusts the hyperparameters as necessary.

[1668] Business decision simulation

[1669] Step 1:

[1670] The terminal provides an interface to the simulation tool for internal users.

[1671] Step 2:

[1672] The user inputs a specific business scenario (e.g., new market entry, product development).

[1673] Step 3:

[1674] The server receives the input scenario and queries the generative AI model.

[1675] Step 4:

[1676] The generative AI model outputs management decisions based on the scenario and sends the results to a server.

[1677] Step 5:

[1678] The server returns the results to the terminal, where the user views the results.

[1679] Management Consulting Platform

[1680] Step 1:

[1681] The terminal provides an interface for the management consulting platform to external users.

[1682] Step 2:

[1683] The user inputs the specific business consultation content (e.g., fundraising methods, human resource management strategies).

[1684] Step 3:

[1685] The server receives the input consultation content and executes a query against the generative AI model.

[1686] Step 4:

[1687] The generative AI model generates appropriate advice and solutions based on the consultation content and sends the results to the server.

[1688] Step 5:

[1689] The server sends advice and solutions to the terminal, which the user then confirms.

[1690] Introducing the Emotion Engine

[1691] Step 1:

[1692] The terminal activates a microphone for capturing the user's voice data and a camera for capturing facial expression data.

[1693] Step 2:

[1694] The server uses a voice analysis module to analyze the captured voice data and recognize the user's emotions (e.g., joy, sadness, anger, surprise).

[1695] Step 3:

[1696] The server uses an image analysis module to analyze the captured facial expression data and recognize the user's emotions.

[1697] Model emotional response

[1698] Step 1:

[1699] The server receives the user's emotional data analyzed by the emotion engine and determines the user's current emotional state.

[1700] Step 2:

[1701] The server adjusts the output of the generative AI model based on the determined emotion data. For example, if the user is nervous, the server adjusts the advice content to be gentler.

[1702] Step 3:

[1703] The generative AI model generates emotion-based tailored advice and business decisions and sends the results to a server.

[1704] Step 4:

[1705] The server sends the adjusted results to the terminal, and the user confirms them.

[1706] Model Update

[1707] Step 1:

[1708] The server collects new data (e.g., the latest interviews or talks) and adds it to the database.

[1709] Step 2:

[1710] The server cleans the new data and standardizes the format.

[1711] Step 3:

[1712] The server uses the updated data to retrain the generative AI model and emotion engine.

[1713] Step 4:

[1714] The server evaluates the performance of the retrained model and releases an update.

[1715] The above are the specific processing steps of the program for the system that combines the emotion engine.

[1716] Example 2

[1717] 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."

[1718] Conventional management support systems have difficulty in fully utilizing the manager's past knowledge and experience, and in providing advice that reflects the user's feelings. As a result, they are unable to provide efficient and appropriate support for management decisions and management consultations, resulting in problems such as lower user satisfaction and success rates.

[1719] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for collecting the manager's past words and actions, books, video recordings, etc., a means for cleaning the collected data and standardizing the format, and a means for training the generative AI model using the preprocessed data. This makes it possible to provide efficient and appropriate management decisions and management consultations while inheriting the manager's knowledge.

[1720] "Past words and actions of management" refers to past words and actions made by management, as well as records of those actions.

[1721] "Books" refers to books written by business managers and books on management.

[1722] "Video recordings" refers to video or recorded media such as speeches or interviews in which the executive appears.

[1723] "Collection methods" refers to the technologies and methods used to obtain relevant data from the internet or physical media.

[1724] "Cleaning and formatting methods" refers to techniques and methods used to remove unnecessary information from collected data and convert it into a format suitable for analysis.

[1725] A "generative AI model" refers to an artificial intelligence model that is trained to reproduce the actions and behavior of managers.

[1726] "Preprocessed data" refers to data that has been cleaned and formatted and is used to train a model.

[1727] "Training means" refers to techniques and methods, including learning algorithms and computing resources, for improving the accuracy of artificial intelligence models.

[1728] "Business situations and scenarios" refer to specific business decision-making situations and strategies that users are considering.

[1729] "Means for generating business decisions" refers to technologies and methods for deriving optimal business decisions based on scenarios input by users using generative AI models.

[1730] "Advice using a generative AI model for business consultations" refers to advice and suggestions generated by AI in response to business-related questions and inquiries from users.

[1731] "Means of updating generative AI models with new data and feedback" refers to techniques and methods that incorporate new data collected and user feedback to improve the accuracy of the model and keep it up to date with the latest information.

[1732] An "emotion-recognizing emotion engine" refers to an artificial intelligence module that analyzes the user's emotional state from their voice and facial expressions.

[1733] "Means for adjusting output" refers to techniques or methods for changing the content or tone of the output of a generative AI model based on the user's emotional state.

[1734] "Speech recognition technology" refers to technology for converting voice data into text data.

[1735] "Optical Character Recognition (OCR)" refers to technology for extracting character data from images and scanned documents.

[1736] "Database" refers to a digital storage system that stores collected data in an organized manner and makes it easily searchable and accessible.

[1737] The present invention relates to a system for making business decisions and providing business advice by combining a generative AI model that has learned the words, actions, and thoughts of business managers with an emotion engine that recognizes the emotions of users. Specific embodiments of this system are described in detail below.

[1738] Data collection

[1739] The server first collects the manager's past statements, actions, books, video recordings, etc. Specifically, it uses web scraping tools and APIs to obtain relevant data from the Internet. It then automatically converts video data into text using voice recognition technology (e.g., Google Speech-to-Text API), and digitizes book and article data using optical character recognition (OCR) technology (e.g., Tesseract OCR).

[1740] Data Preprocessing

[1741] The server then cleans and standardizes the collected data by removing duplicates and noise, tokenizing the text data (splitting words and sentences), and storing it in a consistent format in a database (e.g., MySQL, PostgreSQL).

[1742] Model training

[1743] The server uses the preprocessed data to train a generative AI model. Specifically, it uses a deep learning framework (e.g., TensorFlow or PyTorch) to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning.

[1744] Business decision simulation

[1745] The terminal provides an interface that internal users can access and input specific business situations and scenarios. For example, users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives this, generates business decisions based on the generative AI model, and sends them back to the terminal. The user then refers to the results and makes the actual business decisions.

[1746] Management Consulting Platform

[1747] The terminal provides an interface that external users, such as other managers or entrepreneurs, can access to receive management advice. The user inputs specific consultation content, such as "fundraising strategies" or "recruitment methods." The server receives this information and uses a generative AI model to generate appropriate advice, which is then displayed on the terminal. The user can then use this advice as a reference when making management decisions.

[1748] Introducing the Emotion Engine

[1749] The server is equipped with an emotion engine that recognizes the user's emotions. Specifically, it includes a voice analysis module (e.g., OpenSMILE) that analyzes voice data and an image analysis module (e.g., OpenCV, Dlib) that analyzes facial expression data. The device captures the user's voice and facial expression in real time, and the server analyzes them.

[1750] Model emotional response

[1751] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, the server adjusts the advice content to be gentler. Also, if positive emotions are recognized, the server provides more proactive advice.

[1752] Model Update

[1753] The server periodically updates the generative AI model and emotion engine using new data and feedback. For example, it collects newly acquired data from executive interviews and speeches and adds it to the existing dataset. This allows the generative AI model and emotion engine to always provide decisions based on the latest information.

[1754] Specific examples

[1755] Examples of data collection

[1756] The server downloads past lecture videos of executives from an online platform, converts them into text format using speech recognition technology, and then stores them in a database after a cleaning process.

[1757] Specific examples of business decision simulation

[1758] An in-house project manager accesses the simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, and sends them back to the terminal. The user then formulates an actual market launch plan based on these proposals.

[1759] Examples of management consulting platforms

[1760] External entrepreneurs use their devices to access the business consulting platform and consult about "fundraising strategies for business expansion." The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. Users can follow this advice to smoothly proceed with fundraising activities.

[1761] Examples of emotion engines

[1762] A user accesses the business consulting platform using a terminal and asks a question by voice. The server uses a voice analysis module to analyze the user's emotions and determines that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[1763] By using this system, it becomes possible to make efficient and appropriate management decisions and receive management advice while inheriting the knowledge of management, and it is possible to provide more personalized advice that responds to the user's emotions.

[1764] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1765] Program processing flow

[1766] Step 1: Data collection

[1767] The server collects the manager's past statements, actions, books, and video recordings from the internet. Specifically, it obtains the data using web scraping tools and APIs. Video data is converted into text using voice recognition technology (e.g., Google Speech-to-Text API), and book and article data is digitized using optical character recognition (OCR) technology (e.g., Tesseract OCR). The input is URLs and file paths related to the manager, and the output is a collection of text data.

[1768] Step 2: Data Preprocessing

[1769] The server cleans the collected data and standardizes its format. Specifically, it removes duplicate data and noise. It tokenizes (splits words) and sentences the text data and stores it in a consistent format in a database (e.g., MySQL, PostgreSQL). The input is the text data collected in step 1, and the output is organized data in a unified format.

[1770] Step 3: Model training

[1771] The server uses the preprocessed data to train a generative AI model. A deep learning framework (e.g., TensorFlow or PyTorch) is used to create a model that can reproduce the words, actions, and thoughts of managers. The dataset is divided into a training set and a test set, and the model's performance is improved through reinforcement learning and supervised learning. The input is the preprocessed data from step 2, and the output is a trained generative AI model.

[1772] Step 4: Business decision simulation

[1773] The terminal provides an interface that allows internal users to input specific business situations and scenarios. Users input scenarios such as "entering new markets" or "cost reduction strategies." The server receives the scenarios and simulates business decisions using a generative AI model. The results are sent back to the terminal, and the user refers to them to make the final business decision. The input is the business scenario, and the output is the simulation results.

[1774] Step 5: Management Consulting Platform

[1775] The terminal provides an interface for external users to receive management consultations. Users input specific questions about "fundraising strategies" and "recruitment methods." The server receives the input consultation content, generates appropriate advice using a generative AI model, and displays the output on the terminal. The user uses this advice as a reference when making management decisions. The input is the consultation content, and the output is the generated advice.

[1776] Step 6: Implementing the Emotion Engine

[1777] The device is equipped with a camera and microphone to capture the user's voice and facial expressions in real time. The server analyzes the voice data using a voice analysis module (e.g., OpenSMILE) and the facial expression data using an image analysis module (e.g., OpenCV, Dlib). The input is voice and facial expression data, and the output is the emotional data resulting from the analysis.

[1778] Step 7: Model emotional response

[1779] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion engine. For example, if the user is feeling stressed, it will provide gentler advice, and if positive emotions are recognized, it will provide more proactive advice. The input is emotional data, and the output is the adjusted advice or judgment.

[1780] Step 8: Model Update

[1781] The server periodically updates the generative AI model based on new data and feedback. It adds newly collected interview and lecture data, integrates it with the existing dataset, and retrains the model. The input is the new data and feedback, and the output is an updated generative AI model.

[1782] The above are the specific processing steps of the system program. The operations performed at each step enable the user to make more efficient and appropriate business decisions and receive advice.

[1783] (Application example 2)

[1784] 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."

[1785] In today's business environment, it is difficult to pass on the knowledge and experience of managers to future generations, and there is a need for real-time management decisions and consultations that respond to users' emotions. Monitoring employee mental health is also an important issue, and it is difficult to provide effective countermeasures at the appropriate time. For this reason, there is a need to develop a system that can reproduce the words and actions of managers and provide real-time advice that responds to users' emotions.

[1786] 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 the manager's past words and actions, books, video recordings, etc., means for cleaning the collected data and standardizing the format, means for training the generative AI model using preprocessed data, means for generating management decisions based on specific management situations and scenarios, means for providing advice using the generative AI model in response to a user's management consultation, means for updating the generative AI model using new data and feedback, an emotion recognition engine for recognizing the user's emotions, means for adjusting the output of the generative AI model based on emotion data using the emotion recognition engine, means for analyzing real-time data acquired from sensors, and means for providing appropriate responses and advice in real time using the generative AI model. This enables the provision of personalized advice that corresponds to the user's emotions and real-time mental health monitoring of employees while inheriting the manager's knowledge.

[1787] "Past words and actions of management" refers to records of statements, decisions, actions, etc. made by management in the past.

[1788] "Books" are books written by business managers or management reference works.

[1789] "Video recordings" refers to video data such as videos, interviews, and lectures featuring management.

[1790] "Means of collecting data" refers to methods of obtaining data, such as using web scraping tools, APIs, OCR technology, etc.

[1791] "Means for cleaning data and standardizing the format" refers to methods for removing duplication and noise from collected data, and for tokenizing and segmenting text data.

[1792] "Means for training a generative AI model using preprocessed data" refers to a method for training a generative AI model using cleaned data using a deep learning framework (e.g., TensorFlow or PyTorch).

[1793] A "means for generating business decisions" is a method for using a generative AI model to form business decisions based on specific business situations or scenarios.

[1794] The "means of providing advice" is a method of providing advice to a user's business consultation using a generative AI model.

[1795] A "generative AI model" is a model that uses deep learning technology to reproduce the words, actions, and thoughts of managers and generate management decisions and advice.

[1796] "Feedback-based updating" is a method of retraining a model based on new data and user feedback to improve its performance.

[1797] An "emotion recognition engine" is an engine that uses a voice analysis module and an image analysis module to recognize a user's emotions in real time.

[1798] "Means for adjusting output based on emotional data" refers to a method for adjusting the output results of a generative AI model based on emotional data analyzed by an emotion recognition engine.

[1799] "Means for analyzing real-time data" refers to methods for analyzing data obtained from security cameras and audio sensors in real time.

[1800] "Means for providing appropriate responses and advice in real time" refers to methods for providing appropriate responses and advice on the spot based on data analyzed using a generative AI model.

[1801] The present invention relates to a system that combines an emotion recognition engine and a generative AI model to provide management decisions based on the manager's past words and actions and experience, as well as real-time advice corresponding to the user's emotions. This system aims to make decisions in specific management situations and manage the mental health of employees. A specific embodiment of the present invention is described below.

[1802] Hardware and software used

[1803] 1. Hardware:

[1804] security cameras

[1805] Audio Sensor

[1806] microphone

[1807] GPU-equipped servers

[1808] 2. Software:

[1809] Python

[1810] OpenCV (image analysis library)

[1811] Keras (deep learning framework)

[1812] TensorFlow (deep learning framework)

[1813] SpeechRecognition library (audio analysis)

[1814] GPT-2 model (generative AI model)

[1815] Specific Embodiments of the System

[1816] 1. Data Collection

[1817] The server collects information such as past statements and actions of executives, books, and video recordings from online platforms and internal databases. The collected video data is converted into text using voice recognition technology, and books and articles are converted into digital data using OCR technology.

[1818] 2. Data Preprocessing

[1819] The server cleans the collected data, removes duplicate data and noise, tokenizes the text data, splits it into sentences, and stores it in a unified database.

[1820] 3. Model training

[1821] The server uses the preprocessed data to train a generative AI model. A deep learning framework (TensorFlow or PyTorch) is used to create the generative AI model. The model is divided into a training set and a test set, and its performance is improved through reinforcement learning and supervised learning.

[1822] 4. Emotion recognition

[1823] It analyzes real-time data (audio and video) of employees and users acquired through security cameras and audio sensors, and detects user emotions using an emotion recognition engine. It uses the SpeechRecognition library for audio analysis and OpenCV for image analysis.

[1824] 5. Tuning generative AI models based on emotion data

[1825] The server adjusts the output of the generative AI model based on the user's emotional data analyzed by the emotion recognition engine. For example, if the user is feeling stressed, the generative AI model will be adjusted to output gentle and calming advice.

[1826] 6. Real-time advice

[1827] The server generates appropriate responses and advice based on real-time data analyzed using the generative AI model and displays them on the user's device, making it possible to monitor employee mental health and support management decision-making.

[1828] Specific examples

[1829] The server downloads videos of past executive speeches from an online platform, converts them into text using speech recognition technology, and stores them in a database after a cleaning process. An in-house project manager accesses a simulation tool using a terminal and inputs a scenario called "New Product Market Launch." The server uses a generative AI model to generate risk assessments and strategic proposals based on the specified scenario, which are then sent back to the terminal. A user using the terminal asks a question by voice, and the server uses a speech analysis module to analyze the user's emotions and determine that they are nervous. As a result, the server adjusts the output of the generative AI model to provide calmer, more reassuring advice.

[1830] Prompt Sentence Examples

[1831] "Generate examples of measures to take when an employee is under stress."

[1832] A system based on this format will utilize the knowledge of management, provide advice that responds to the user's emotions, and enable real-time mental health monitoring of employees.

[1833] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1834] Step 1:

[1835] The server collects information about the manager's past statements, actions, books, video recordings, etc. from online platforms and internal databases. Specifically, it acquires data using web scraping tools, APIs, and OCR technology. The inputs are the URL of the online platform and a database query, and the acquired data (text, audio, video) is output.

[1836] Step 2:

[1837] The server cleans the collected data by removing duplicates, removing noise, tokenizing text data, and segmenting sentences. The input is the collected raw data, and the output is cleaned, consistent data.

[1838] Step 3:

[1839] The server uses the preprocessed data to train a generative AI model. Specifically, it builds the model using a deep learning framework (TensorFlow or PyTorch) and performs reinforcement learning or supervised learning using a training set and a test set. It takes the preprocessed data as input and outputs a trained generative AI model.

[1840] Step 4:

[1841] A user accesses the business decision simulation tool using a terminal and inputs a specific business situation or scenario, such as "entering a new market" or "cost reduction strategy." The scenario information is input, and a simulation request is sent to the server.

[1842] Step 5:

[1843] The server uses a generative AI model to generate business decisions based on the specified scenario. Specifically, it analyzes the input scenario information and generates business decisions and strategic proposals. A scenario request is input, and the generated business decisions are output and sent to the terminal.

[1844] Step 6:

[1845] A user using a terminal accesses the business consulting platform and inputs a specific consultation content (e.g., "Fundraising strategy for business expansion"). With the consultation content as input, a consultation request is sent to the server.

[1846] Step 7:

[1847] The server uses a generative AI model to generate specific advice based on past successes and failures, and displays it on the device. A consultation request is input, and the generated advice is output.

[1848] Step 8:

[1849] The device uses security cameras and audio sensors to capture real-time audio and video data to analyze the user's emotions. An emotion recognition engine is used to analyze the audio and facial expression data to identify the user's...

Claims

1. A means of collecting the manager's past words and actions, books, video recordings, etc. A means of cleaning the collected data and standardizing its format; means for training a generative AI model using the preprocessed data; A means of generating business decisions based on specific business situations and scenarios; A means for providing advice using a generative AI model in response to a user's business consultation; A system that includes a means to update a generative AI model with new data and feedback.

2. 10. The system of claim 1, further comprising means for storing the collected data in a database.

3. 2. The system according to claim 1, further comprising means for transmitting the results of management decisions and advice to a user terminal and displaying them.

Images