system

Abstract

We provide the system. [Solution] A means of collecting the careers, past decisions, statements, and related literature of executives and generating them as a unified dataset, A method for training a machine learning model based on preprocessed data and generating a digital twin of executives, A means of analyzing data and scenarios related to new management decisions and performing simulations, A system that includes means for visually displaying the results of a simulation to the user.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, and includes steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance as a response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In enterprises, there is a current situation where it is difficult to utilize the experience and knowledge of entrepreneurs and senior executives in future business decisions. In particular, it is not easy to inherit the decision-making process and unique business strategies of entrepreneurs for future generations. Also, in new business decisions and strategic determinations, there is a lack of means to reflect the insights of these executives in real time. As a result, the problem is that efficient and effective business decisions are not made.

Means for Solving the Problems

[0005] This invention provides a means for collecting executives' career histories, past decisions, statements, and related literature, and generating them as a centralized dataset. Furthermore, by incorporating means for training machine learning models based on preprocessed data and generating digital twins of executives, these insights are digitized and stored. In addition, means for analyzing data and scenarios related to new management decisions, running simulations, and visually displaying the simulation results to the user enable executive insights to be reflected in management decisions in real time. This allows companies to formulate strategies from a long-term perspective and make efficient and effective management decisions.

[0006] The term "executive" refers to a high-ranking manager who makes decisions within a company or organization.

[0007] "Career history" refers to the executive's work experience, educational background, past achievements, and other historical information.

[0008] "Decisions" refers to records of decisions, judgments, and transactions made by executives in the course of their duties.

[0009] "Records of statements" refers to information that records the content of statements made by executives in meetings or public forums.

[0010] "Literature" refers to research reports, papers, books, etc., related to management or the industry.

[0011] A "dataset" refers to a collection of data that has been gathered and organized for a specific purpose.

[0012] "Preprocessing" refers to the process of preparing collected data into a format suitable for analysis and machine learning.

[0013] A "machine learning model" refers to an algorithm that learns patterns based on large amounts of data and uses those results to perform predictions and classifications.

[0014] A "digital twin" refers to a digital replica that mimics real-world events in real time.

[0015] "Scenario" refers to a set of input data assuming specific business judgments or strategies.

[0016] "Simulation" refers to a process of simulating what results actual business judgments or strategies will bring.

[0017] "User" refers to a person who uses this system to make business judgments or simulations.

[0018] "Visually display" refers to presenting information expressed in numerical values or words to the user in the form of graphs, charts, dashboards, etc.

Brief Description of the Drawings

[0019] [Figure 1] It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which multiple emotions are mapped. [Figure 10]Shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Mode for Carrying Out the Invention

[0020] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0021] First, the terms used in the following description will be explained.

[0022] In the following embodiments, a numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0023] In the following embodiments, a numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0024] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0025] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0026] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0027] [First Embodiment]

[0028] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0029] As shown in Figure 1, the 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.

[0030] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0031] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0032] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and 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.

[0033] 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 perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0034] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0035] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] As shown in Figure 2, in the data processing device 12, a specific processing 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" related to the technology of this 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 according to the specific processing program 56 executed on the RAM 30.

[0037] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0038] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0039] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0040] This invention is a system for digitizing executives' expertise and supporting management decisions, and a specific embodiment thereof is shown below.

[0041] System Configuration

[0042] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[0043] Program processing

[0044] Data collection and transmission

[0045] User

[0046] Users input data such as executives' resumes, past decisions, statements, and relevant literature into the terminal.

[0047] Users collect information from external data sources and upload it to their devices.

[0048] terminal

[0049] The terminal centrally manages the input data and builds a dataset for transmission to the server.

[0050] The terminal standardizes the data format and corrects inconsistencies.

[0051] server

[0052] The server receives data sent from the terminal and stores it in the database.

[0053] Data preprocessing and model generation

[0054] server

[0055] The server preprocesses the received data using natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[0056] The server trains a machine learning model based on pre-processed data to generate a digital twin of the executives.

[0057] The server saves the generated digital twin model to a database.

[0058] Simulation execution and result display

[0059] User

[0060] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[0061] terminal

[0062] The terminal sends the entered data to the server.

[0063] server

[0064] The server runs simulations using a digital twin model and generates prediction results. These simulation results include risk assessments, expected profits, and optimal strategies.

[0065] The server exports the simulation results as a dataset and sends it to the terminal.

[0066] terminal

[0067] The terminal visually displays the simulation results received from the server. This can be done using formats such as graphs, charts, and dashboards.

[0068] Specific example

[0069] Example 1: Launching a new product into market

[0070] User

[0071] The user enters information about the market launch of a new smart device model into the terminal (e.g., key product features, target market, sales budget).

[0072] terminal

[0073] The terminal normalizes the input data and sends it to the server as a dataset.

[0074] server

[0075] The server runs simulations using digital twin models of executives, generating information such as projected sales, risk factors, and marketing strategies.

[0076] The server sends the simulation results to the terminal.

[0077] terminal

[0078] The terminal visually displays the simulation results received from the server to the user.

[0079] Example 2: Considering a business acquisition

[0080] User

[0081] The user enters information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal.

[0082] terminal

[0083] The terminal sends the input data to the server as a dataset.

[0084] server

[0085] The server runs simulations using digital twin models of executives. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[0086] The server sends the simulation results to the terminal.

[0087] terminal

[0088] The terminal visually displays the simulation results received from the server to the user.

[0089] This system digitizes the knowledge and experience of executives and utilizes it in real time for management decisions, thereby supporting companies in effectively formulating long-term strategies.

[0090] The following describes the processing flow.

[0091] Step 1:

[0092] User

[0093] Users input executives' career histories, past decisions, statements, and relevant documents into the terminal. This includes text data and scanned documents.

[0094] Step 2:

[0095] terminal

[0096] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[0097] Step 3:

[0098] terminal

[0099] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[0100] Step 4:

[0101] terminal

[0102] The terminal sends the centralized dataset to the server. During this process, it detects data transfer errors and attempts to resend the data if necessary.

[0103] Step 5:

[0104] server

[0105] The server receives data sent from the terminal and stores it in a database. This database is used for subsequent processing.

[0106] Step 6:

[0107] server

[0108] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[0109] Step 7:

[0110] server

[0111] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning.

[0112] Step 8:

[0113] server

[0114] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[0115] Step 9:

[0116] server

[0117] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[0118] Step 10:

[0119] User

[0120] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[0121] Step 11:

[0122] terminal

[0123] The terminal sends the new data and scenario entered by the user to the server.

[0124] Step 12:

[0125] server

[0126] The server analyzes new data using a digital twin model and runs simulations. The simulation results include risk assessments, expected profits, and optimal strategies.

[0127] Step 13:

[0128] server

[0129] The server exports the simulation results as a dataset and sends it to the terminal.

[0130] Step 14:

[0131] terminal

[0132] The terminal displays the simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards, etc.).

[0133] Step 15:

[0134] User

[0135] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[0136] (Example 1)

[0137] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0138] Traditional management decision-making systems struggle to digitize the knowledge and experience of executives and utilize it in real time, making it difficult to improve the quality of management decisions. Furthermore, there is a need for a system that can centrally manage and effectively analyze data on executives' past decisions and statements.

[0139] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0140] This invention includes a server that collects the career history, past decisions, statements, and related literature of executives and generates them as a centralized dataset; a server that normalizes, tokenizes, removes stop words, and stems the data using natural language processing techniques based on the preprocessed data; and a server that trains a machine learning model based on the preprocessed data to generate a digital twin of the executives. This makes it possible to digitize the knowledge and experience of executives and utilize it in management decisions in real time.

[0141] "Executives" refer to individuals who make important decisions within a company or organization, and primarily include management and executives.

[0142] "Career history" refers to the historical information of an employee, including their past job duties, achievements, and educational background.

[0143] "Past decisions" refers to specific decisions, judgments, or statements of intent made by an executive in the past.

[0144] "Record of statements" refers to data in text format that records statements and comments made by officials in the past.

[0145] "Literature" refers to materials and information sources such as books, papers, and business reports related to the person in charge.

[0146] A "dataset" refers to a collection of data that has been centrally managed and organized for use in analysis and model training.

[0147] "Natural language processing technology" refers to techniques for converting text data into a format that is easily understood by machines, and includes normalization, tokenization, stop word removal, stemming, and other methods.

[0148] "Normalization" refers to the process of handling character size, spaces, and special characters in order to unify the format of data.

[0149] "Tokenization" is the process of dividing text into words and phrases, which means breaking down a sentence into units that are easier to analyze.

[0150] "Stop word removal" refers to the process of excluding certain common words (such as "a" or "the") from the analysis.

[0151] "Stemming" is the process of extracting only the stem portion of a word, and it refers to unifying the inflected forms of the word.

[0152] A "machine learning model" refers to an algorithm that learns patterns and regularities based on data and uses that knowledge to make predictions and classifications on new data.

[0153] A "digital twin" is a digital representation that mimics the knowledge and judgment of real-world officials, and is used for simulations and predictions.

[0154] A "scenario" refers to a specific situation or plan related to a new management decision.

[0155] "Simulation" refers to using digital twin models to make predictions and analyses about new situations and scenarios.

[0156] "Visualization" refers to techniques that visually display simulation results using graphs, charts, and other methods to make them easier to understand.

[0157] This invention is a system for digitizing the knowledge of executives and supporting management decisions. Its specific embodiments are as follows:

[0158] System Configuration

[0159] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[0160] Data collection and transmission

[0161] User

[0162] Users input data such as the career history of executives, past decisions, records of statements, and relevant literature into spreadsheets or dedicated forms. They also collect information from external data sources using APIs and scraping tools and upload it to their devices.

[0163] terminal

[0164] The terminal converts the input data into a CSV file and builds a dataset for centralized management. It standardizes the data format using Excel or Google Sheets, detects and corrects inconsistent data, and sends the corrected data to the server.

[0165] server

[0166] The server receives data sent from the terminal and saves it to a dedicated database (e.g., PostgreSQL). During this process, it checks the data's integrity to ensure there is no invalid data.

[0167] Data preprocessing

[0168] server

[0169] The server uses natural language processing techniques (e.g., Python's NLP library spaCy) to normalize, tokenize, remove stop words from, and stem the stored data. For example, all characters are converted to lowercase, unnecessary spaces and special characters are removed, and then tokenization is performed. This divides the data into units of words and phrases, common words (stop words) are removed, and stemming is performed to extract only the word stems.

[0170] Generation of a digital twin model

[0171] server

[0172] The server trains a digital twin model using machine learning algorithms (e.g., TENSORFLOW®, scikit-learn) based on preprocessed data. For example, it uses random forests or neural networks to generate a digital twin, which is a digital representation that mimics the knowledge of the person in charge. The trained model is stored in a dedicated database for performance evaluation and optimization.

[0173] Running the simulation and displaying the results.

[0174] User

[0175] The user inputs data and scenarios related to new business decisions into the terminal. For example, they might input a specific scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[0176] terminal

[0177] The terminal converts the entered scenario data into JSON format and sends it to the server.

[0178] server

[0179] The server uses a digital twin model to perform scenario-based simulations and generates predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising" and send them to the terminal.

[0180] terminal

[0181] The terminal visually displays the received simulation results. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[0182] Specific example

[0183] Example 1: Launching a new product into market

[0184] User

[0185] The user inputs information about the market launch of a new smart device model into the terminal. For example, they might input, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[0186] terminal

[0187] The terminal converts the input data into JSON format and sends it to the server.

[0188] server

[0189] The server runs a simulation using a digital twin model, generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," and sends them to the terminal.

[0190] terminal

[0191] The terminal displays the received results in graphs. Sales forecasts are shown as bar graphs, risk assessments as heatmaps, and recommended strategies as lists.

[0192] Example 2: Considering a business acquisition

[0193] User

[0194] The user inputs the following data into the terminal: "Company Z is being considered for acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[0195] terminal

[0196] The terminal converts this data into a CSV file and sends it to the server.

[0197] server

[0198] The server runs a simulation using a digital twin model, generates results such as "Risk assessment: High, Expected return: 2 billion yen, Recommended strategy: Enhanced due diligence," and sends them to the terminal.

[0199] terminal

[0200] The terminal displays the received results in charts. Risk assessment is shown as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[0201] This will enable the digitization of executives' knowledge and experience, allowing them to be used in real-time management decisions. The system will support companies in effectively formulating long-term strategies.

[0202] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0203] Step 1: Data collection and input

[0204] User

[0205] Users input data such as the career history, past decisions, statements, and relevant literature of executives into spreadsheets or dedicated forms. For example, a user might input "Executive A's career history, past decisions, and relevant literature" into a spreadsheet. They can also collect information from external data sources using APIs or scraping tools and upload it to their device. The input data is saved in CSV or Excel format.

[0206] terminal

[0207] The terminal receives the input data, converts it into a CSV file, and builds a centrally managed dataset. For example, it standardizes the data format using Excel or Google Sheets, detects inconsistent data, and corrects it. The corrected data is prepared as a temporary storage file.

[0208] Input: Records of the person's career history, past decisions, statements, and relevant literature (CSV format file)

[0209] Output: Centrally managed dataset (CSV format file)

[0210] Step 2: Send

[0211] terminal

[0212] The terminal sends a centrally managed dataset to the server. During this process, the integrity of the files is checked, and retransmission is performed if necessary. For example, a data transmission API is used to send the dataset to the server.

[0213] Input: Centrally managed dataset (CSV format file)

[0214] Output: Request to send data to the server

[0215] server

[0216] The server receives data sent from the terminal and stores it in a database. For example, it might store it in a PostgreSQL database and perform data integrity checks.

[0217] Input: Data transmission request

[0218] Output: Data stored in the database

[0219] Step 3: Data Preprocessing

[0220] server

[0221] The server preprocesses the stored data using natural language processing techniques. First, it performs text normalization, converting all characters to lowercase and removing unnecessary spaces and special characters. Next, it performs tokenization, dividing the data into units of words and phrases. Furthermore, it removes stop words and extracts the stem portion of words using stemming.

[0222] Input: Data stored in the database

[0223] Output: Preprocessed text data

[0224] Specific operation: Use the Python spaCy library to perform text normalization, tokenization, stop word removal, and stemming.

[0225] Step 4: Generating a Digital Twin Model

[0226] server

[0227] The server trains a digital twin model using machine learning algorithms based on pre-processed data. Specifically, it uses random forest and neural network algorithms to generate a digital twin that mimics the knowledge of the executives. The trained model undergoes performance evaluation and optimization before being stored in a database.

[0228] Input: Preprocessed text data

[0229] Output: Trained digital twin model

[0230] Specific actions: Train the model using TensorFlow or scikit-learn.

[0231] Step 5: Run the simulation

[0232] User

[0233] The user inputs a scenario related to a new business decision into the terminal. For example, they might input a scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[0234] terminal

[0235] The terminal converts the entered scenario data into JSON format and sends it to the server.

[0236] Input: Scenario (text format)

[0237] Output: Scenario data in JSON format

[0238] server

[0239] The server uses a digital twin model to perform scenario-based simulations and generate predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising."

[0240] Input: Scenario data in JSON format

[0241] Output: Simulation results (JSON format)

[0242] Step 6: Displaying the simulation results

[0243] terminal

[0244] The terminal visually displays the simulation results received from the server. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[0245] Input: Simulation results (JSON format)

[0246] Output: Visualized simulation results (graphs, heatmaps, lists)

[0247] Specific example

[0248] Example 1: Launching a new product into market

[0249] User

[0250] The user enters the following information into the terminal: "New product Y to be launched, target market: men in their 20s, sales budget: 30 million yen."

[0251] terminal

[0252] The device normalizes this information, converts it to JSON format, and sends it to the server.

[0253] server

[0254] The server runs a simulation using a digital twin model and generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," which it then sends to the terminal.

[0255] terminal

[0256] The device visually displays the results, showing sales forecasts as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[0257] Example 2: Considering a business acquisition

[0258] User

[0259] The user enters the following information into the terminal: "Company Z is considering acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[0260] terminal

[0261] The terminal normalizes this information, converts it to CSV format, and sends it to the server.

[0262] server

[0263] The server runs a simulation using a digital twin model and generates results such as "Risk Assessment: High, Expected Return: 2 billion yen, Recommended Strategy: Enhanced Due Diligence," which it then sends to the terminal.

[0264] terminal

[0265] The device visually displays the results, showing risk assessments as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[0266] Following these specific processing steps, the system can digitize the knowledge and experience of executives and utilize it in real time for management decisions.

[0267] (Application Example 1)

[0268] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0269] In modern corporate management, there is a need to leverage the insights of executives and management in real time to effectively support management decisions. Furthermore, it is crucial to efficiently collect operational data from machinery and equipment used in factories and formulate optimal operational plans. However, systems to address these challenges are still not adequately developed, resulting in insufficient improvements in the quality of management decisions and the efficiency of factory operations.

[0270] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0271] This invention includes a server that collects executives' careers, past decisions, records of statements, and related literature and generates them as a centralized dataset; a server that trains a machine learning model based on the pre-processed data to generate a digital twin of the executives; a server that collects operational data of machinery and equipment used in the factory and generates an optimized operational plan; a server that analyzes data and scenarios related to new management decisions and runs simulations; and a server that visually displays the results of the simulations to the user. This enables the real-time utilization of management insights to support management decisions and further improves and optimizes the efficiency of factory operations.

[0272] An "executive" is a person who is in a position to make important decisions within a company or organization and who possesses extensive knowledge and experience in management.

[0273] "Background" refers to information that shows what kind of jobs and roles a person has experienced in the past.

[0274] A "record of statements" is data that compiles and saves statements and comments made by a specific person in the past.

[0275] "Related literature" refers to documentary materials such as books, articles, and reports related to the person or event in question.

[0276] The "integrated dataset" is a collection of data integrated from multiple different information sources and organized in a form that is easy to manage.

[0277] The "machine learning model" is a data analysis algorithm for making predictions and classifications based on the collected data.

[0278] The "digital twin" is a virtual replica of a real object, for example, one that emulates the knowledge and experience of executives as data.

[0279] The "mechanical devices used in the factory" refers to the machines and equipment used in the manufacturing and production processes.

[0280] The "operation data" is specific numerical data related to the operation status, such as the operation status of mechanical devices, production volume, and the number of error occurrences.

[0281] The "optimized operation plan" is a plan formulated aiming at the efficient and effective operation of mechanical devices.

[0282] The "data related to new business decisions" is data indicating information and scenarios about new business strategies, investment plans, etc.

[0283] The "simulation" is a test or experiment conducted by simulating the real situation, and it is a method for making predictions and evaluations under specific conditions.

[0284] The "means for visually displaying" refers to a method of displaying data and simulation results in a form that is easy for users to understand, including graphs, charts, dashboards, etc.

[0285] The present invention is a system for digitizing the knowledge of executives and supporting business decisions. The system consists of three components: a terminal, a server, and a user. The specific embodiments will be described below.

[0286] System Configuration

[0287] The following hardware and software are mainly used in this system:

[0288] Hardware: Smartphones, tablets (terminals), servers

[0289] Software: Python, Pandas, Scikit-learn, Matplotlib

[0290] Program Processing

[0291] Data Collection and Transmission

[0292] Users input the history of officers, past decisions, records of speeches, relevant documents, etc. into the terminal using smartphones or tablets. In addition, users also input the operation data of mechanical devices used in the factory (e.g., operating hours, number of error occurrences, production quantity). The terminal normalizes the input data and transmits it to the server as a unified dataset.

[0293] Data Preprocessing and Model Generation

[0294] The server receives the data transmitted from the terminal and performs preprocessing such as text normalization, tokenization, and standardization. Then, the server trains a machine learning model based on the preprocessed data and generates a digital twin of the officer. In addition, the server uses the operation data of mechanical devices used in the factory to generate an optimized operation plan and saves each model.

[0295] Simulation Execution and Result Display

[0296] Users input data and scenarios related to new business decisions into a terminal. Examples include scenarios for new product launches or business acquisitions, annual production targets, and maintenance schedules. The terminal sends this data to a server. The server runs simulations using a digital twin model and operational data model to generate forecast results. These forecast results include risk assessments, expected profits, and production plans. The server sends the simulation results back to the terminal, which displays them visually. Display formats include graphs, charts, and dashboards.

[0297] Specific example: Optimizing the annual operational plan for factory robots.

[0298] The user inputs annual operational data (e.g., monthly production targets, maintenance schedules for various machines, etc.). The following prompts are used as an example:

[0299] Example of an input prompt:

[0300] Please enter the following operational data:

[0301] Months: January, February, March...December

[0302] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[0303] Maintenance scheduled for: January 15th, February 20th...

[0304] Starting the simulation.

[0305] This allows users to input necessary data through their terminals, and the server to run simulations using operational data models, generating and presenting optimal operational plans.

[0306] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0307] Step 1:

[0308] The user inputs into the terminal the resume of officers, past decisions, records of remarks, relevant literature, and operation data of mechanical devices used in the factory. The input data includes officers' attribute information (e.g., name, past positions), details of remarks, citations of literature, operating hours, number of error occurrences, production quantity, etc. The terminal collects these data centrally, normalizes the data as necessary, and corrects inconsistencies.

[0309] Step 2:

[0310] The terminal constructs the collected and normalized data as a dataset and transmits it to the server. At this time, the data is converted into a unified format such as a CSV file or JSON format. The destination server receives this data and stores it in the database.

[0311] Step 3:

[0312] The server preprocesses the received data. This includes natural language processing (NLP) techniques such as text normalization, tokenization, and stop word removal. Also, for numerical data, preprocessing such as standardization and normalization is performed. The preprocessed data is converted into an analyzable form and proceeds to the next processing.

[0313] Step 4:

[0314] The server trains a machine learning model using the preprocessed data. This includes a digital twin model of officers and a model for generating an optimal operation plan for mechanical devices used in the factory. These models are trained using random forest regression and deep learning techniques. The generated models are stored in the server.

[0315] Step 5:

[0316] The user inputs data and scenarios related to new business decisions into the terminal. Specifically, they provide scenario information such as new product launches, business acquisitions, and new production targets. The following prompts are used:

[0317] Example of an input prompt:

[0318] Please enter the following operational data:

[0319] Months: January, February, March...December

[0320] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[0321] Maintenance scheduled for: January 15th, February 20th...

[0322] Starting the simulation.

[0323] Step 6:

[0324] The terminal sends the new data and scenario entered by the user to the server. The server receives this data, performs preprocessing and standardization, and then runs a simulation using the stored digital twin model and operational planning model.

[0325] Step 7:

[0326] The server generates the results of the executed simulation. These predictions include risk assessments, expected profits, and optimized production plans. The simulation results are stored as a dataset on the server and sent to the terminal.

[0327] Step 8:

[0328] The terminal visually displays simulation results received from the server to the user. This includes graphs, charts, and dashboards. Users can use this information to make specific business decisions and operational plan decisions.

[0329] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0330] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions, and a specific embodiment thereof is shown below.

[0331] System Configuration

[0332] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[0333] Program processing

[0334] Data collection and transmission

[0335] User

[0336] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Furthermore, users input their own emotional state or have it automatically recognized by an emotion engine.

[0337] terminal

[0338] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted for subsequent processing. The device also sends emotion recognition requests to the emotion engine.

[0339] Emotional Engine

[0340] The emotion engine analyzes the user's emotions using user input data or facial and speech recognition technologies. The analysis results are sent to a server for later use in simulations.

[0341] server

[0342] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database.

[0343] Data preprocessing and model generation

[0344] server

[0345] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[0346] server

[0347] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[0348] server

[0349] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[0350] server

[0351] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[0352] Simulation execution and result display

[0353] User

[0354] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[0355] terminal

[0356] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[0357] server

[0358] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[0359] server

[0360] The server exports the simulation results as a dataset and sends it to the terminal.

[0361] terminal

[0362] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[0363] Specific example

[0364] Example 1: Launching a new product into market

[0365] User

[0366] The user inputs information about the market launch of the new smart device (e.g., key product features, target market, sales budget) into the device. Emotional states are also input simultaneously or recognized by the emotion engine.

[0367] terminal

[0368] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[0369] server

[0370] The server runs simulations using digital twin models of executives and data from the emotion engine. It generates information such as predicted sales, risk factors, and marketing strategies.

[0371] server

[0372] The server sends the simulation results to the terminal.

[0373] terminal

[0374] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[0375] Example 2: Considering a business acquisition

[0376] User

[0377] The user inputs information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal. Emotional states are also input simultaneously or recognized by an emotion engine.

[0378] terminal

[0379] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[0380] server

[0381] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[0382] server

[0383] The server sends the simulation results to the terminal.

[0384] terminal

[0385] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[0386] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[0387] The following describes the processing flow.

[0388] Step 1:

[0389] User

[0390] Users input executives' career histories, past decisions, statements, and relevant literature into the terminal. If emotional states are also to be entered, they can be done using text or a dedicated emotion input interface.

[0391] Step 2:

[0392] terminal

[0393] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[0394] Step 3:

[0395] terminal

[0396] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[0397] Step 4:

[0398] terminal

[0399] The terminal sends the input data to the server. At the same time, it sends an emotion recognition request to the emotion engine.

[0400] Step 5:

[0401] Emotional Engine

[0402] The emotion engine analyzes the user's emotions using user input data, facial recognition, and speech recognition technology. The analysis results are then sent to the server.

[0403] Step 6:

[0404] server

[0405] The server receives data sent from the terminal and the emotion engine and stores it in a database. This database is used for subsequent processing.

[0406] Step 7:

[0407] server

[0408] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[0409] Step 8:

[0410] server

[0411] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[0412] Step 9:

[0413] server

[0414] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[0415] Step 10:

[0416] server

[0417] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[0418] Step 11:

[0419] User

[0420] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[0421] Step 12:

[0422] terminal

[0423] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[0424] Step 13:

[0425] server

[0426] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[0427] Step 14:

[0428] server

[0429] The server exports the simulation results as a dataset and sends it to the terminal.

[0430] Step 15:

[0431] terminal

[0432] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[0433] Step 16:

[0434] User

[0435] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[0436] (Example 2)

[0437] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0438] In recent years, it has become increasingly important for companies to consider not only the experience and knowledge of executives but also the emotions and subjectivity of users when making management decisions. However, there is a lack of systems that can properly centralize, analyze, and visually display the results of simulations based on this information. As a result, the management decision-making process is not conducted efficiently and effectively, leading to a decline in the quality of decision-making in corporate operations.

[0439] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0440] In this invention, the server includes means for inputting or recognizing the user's emotional state, means for collecting the executive's career history, past decisions, records of statements, and related literature, and for formatting and storing them as a centralized dataset, means for training a machine learning model based on the pre-processed data and generating a digital twin of the executive, means for analyzing data and scenarios related to new management decisions while considering the user's emotional state, and for running simulations, and means for exporting the simulation results as a dataset and further displaying it visually. This makes it possible to comprehensively manage and analyze diverse information and support management decisions that reflect emotions.

[0441] "User emotional state" refers to the user's emotions and psychological state, and includes information input through the human-machine interface and emotional information recognized using an emotion engine.

[0442] A "centralized dataset" is a collection of data that integrates and manages executives' career histories, past decisions, statements, and related literature, and then formats and stores them in a format suitable for analysis and processing.

[0443] "Preprocessed data" refers to data that has been transformed into a format suitable for training machine learning models by applying processing such as normalization, tokenization, stop word removal, and stemming to the input data using techniques such as natural language processing.

[0444] A "machine learning model" is an algorithm and system that learns patterns from data and uses that knowledge to make predictions, classifications, and identifications.

[0445] A "digital twin" is a digital twin model created based on machine learning models to reproduce physical executives, their actions, and decision-making processes in a virtual environment.

[0446] A "scenario" refers to a specific combination of circumstances or conditions that should be considered for business decisions, and specifically describes case studies such as the launch of a new product or the acquisition of a business.

[0447] A "simulation" is a computational process that involves trial and error in a virtual environment before actually performing an action, and uses the results to derive risk assessments and optimal strategies.

[0448] "Visual display methods" refer to ways of providing data and analysis results to users in an intuitive and easy-to-understand manner using formats such as graphs, charts, and dashboards.

[0449] A "generative AI model" is an artificial intelligence model that learns patterns and characteristics of data and makes predictions and suggestions based on new data.

[0450] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions. The system is configured as follows:

[0451] System Configuration

[0452] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[0453] Program processing

[0454] Data collection and transmission

[0455] User

[0456] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Users also input their own emotional state or undergo automatic recognition by an emotion engine.

[0457] terminal

[0458] The terminal centrally manages user input data and builds it into a dataset. The data is formatted into an appropriate format. Furthermore, the terminal sends emotion recognition requests to the emotion engine. A typical PC or mobile device can be used as the terminal.

[0459] Emotional Engine

[0460] The emotion engine analyzes user emotions using user input data, facial recognition, and speech recognition technologies. Specifically, it performs facial recognition using Python and the OpenCV library, and speech analysis using Praat. The analysis results are sent to the server in real time.

[0461] server

[0462] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL®).

[0463] Data preprocessing and model generation

[0464] server

[0465] The server preprocesses the received data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming. The preprocessed data is then used to train a machine learning model using TensorFlow or PyTorch. Furthermore, the user's emotional state is also utilized as a parameter in the model.

[0466] server

[0467] The server optimizes the model parameters using a dataset to generate a digital twin for each executive. During this process, the executives' behavioral patterns and decision-making processes are learned. The optimized digital twin model is then stored in a database.

[0468] Simulation execution and result display

[0469] User

[0470] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[0471] terminal

[0472] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[0473] server

[0474] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[0475] server

[0476] The server exports the simulation results as a dataset and sends those results to the terminal.

[0477] terminal

[0478] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (graphs, charts, dashboards, etc.). Data visualization libraries such as D3.js and Chart.js are used for display. In addition, feedback information from the emotion engine is also displayed.

[0479] Specific example

[0480] Example 1: Launching a new product into market

[0481] User

[0482] Users input information about the market launch of a new smart device (key product features, target market, sales budget) into the device. Their emotional state is also input simultaneously or recognized by an emotion engine.

[0483] terminal

[0484] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[0485] server

[0486] The server runs simulations using executive digital twin models and emotion engine data, generating information on predicted sales, risk factors, and marketing strategies.

[0487] server

[0488] The server sends the simulation results to the terminal.

[0489] terminal

[0490] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[0491] Example 2: Considering a business acquisition

[0492] User

[0493] The user inputs information about the target company (financial status, synergy effects, acquisition price) into the terminal. Their emotional state is also entered simultaneously or recognized by an emotion engine.

[0494] terminal

[0495] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[0496] server

[0497] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results for acquisition risk assessment, expected returns, and optimal strategies.

[0498] server

[0499] The server sends the simulation results to the terminal.

[0500] terminal

[0501] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[0502] Example of a prompt

[0503] 1. "A user is considering launching a new product. Use an emotion engine to analyze the user's emotional state and display simulation results to support their decision-making regarding market launch."

[0504] 2. "A user is considering acquiring a certain company. Based on the target company's data and the results of the sentiment engine, perform a simulation of acquisition risk and expected return, and propose the optimal strategy."

[0505] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[0506] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0507] The processing flow of a system program is divided into processing steps.

[0508] Step 1:

[0509] User

[0510] Users input data such as executives' resumes, past decisions, statements, and relevant literature into the terminal. This input includes text data and scanned documents. In addition, users can manually input their own emotional state or have it automatically recognized by an emotion engine.

[0511] Input: Executive data, literature data, emotional state

[0512] Output: Centralized dataset

[0513] Step 2:

[0514] terminal

[0515] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted into an appropriate format. The device temporarily stores the formatted data and sends an emotion recognition request to the emotion engine.

[0516] Input: User data, emotional state

[0517] Data processing: Data format conversion, data formatting.

[0518] Output: Formatted dataset, sentiment recognition requests

[0519] Step 3:

[0520] Emotional Engine

[0521] The emotion engine analyzes user input data and video and audio data obtained from the device's camera and microphone. It uses facial recognition technology and voice analysis (voice feature extraction and emotion classification) to quantify the user's emotional state. The analysis results are transmitted to the server in real time.

[0522] Input: Video data, audio data

[0523] Data processing: facial recognition, voice analysis

[0524] Output: Sentiment analysis results

[0525] Step 4:

[0526] server

[0527] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL).

[0528] Input: Formatted dataset, sentiment analysis results

[0529] Data processing: Data reception, data storage

[0530] Output: Saved database

[0531] Step 5:

[0532] server

[0533] The server preprocesses the stored data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming.

[0534] Input: Saved data

[0535] Data processing: Text preprocessing (normalization, tokenization, stop word removal, stemming)

[0536] Output: Preprocessed data

[0537] Step 6:

[0538] server

[0539] The server trains machine learning models using TensorFlow or PyTorch based on preprocessed data. Furthermore, it utilizes the user's emotional state as a parameter in the model.

[0540] Input: Preprocessed data, emotional state

[0541] Data processing: Machine learning model training

[0542] Output: Trained machine learning model

[0543] Step 7:

[0544] server

[0545] The server optimizes the parameters of the digital twin model using executive data. In this process, the executives' behavioral patterns and decision-making processes are learned.

[0546] Input: Executive data, machine learning model

[0547] Data Calculation: Model Parameter Optimization

[0548] Output: Optimized digital twin model

[0549] Step 8:

[0550] server

[0551] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[0552] Input: Optimized digital twin model

[0553] Data processing: Data storage

[0554] Output: Saved digital twin model

[0555] Step 9:

[0556] User

[0557] Users input data and scenarios related to new business decisions into the terminal. Specific examples include information on new product launches and business acquisitions. Emotional states are either entered sequentially or automatically updated by the emotion engine.

[0558] Input: New scenario data, emotional state

[0559] Output: Input scenario data

[0560] Step 10:

[0561] terminal

[0562] The device sends new scenario data and sentiment analysis results to the server.

[0563] Input: Input scenario data, sentiment analysis results

[0564] Data processing: Data transmission

[0565] Output: Sent data

[0566] Step 11:

[0567] server

[0568] The server uses information from the digital twin model and the emotion engine to perform simulations and generate prediction results. Specific examples include risk assessment, expected profit, and optimal strategy.

[0569] Input: Digital twin model, scenario data, sentiment analysis results

[0570] Data calculation: Simulation execution

[0571] Output: Simulation results

[0572] Step 12:

[0573] server

[0574] The server exports the simulation results as a dataset and sends it to the terminal.

[0575] Input: Simulation results

[0576] Data processing: Data export, data transmission

[0577] Output: Sent simulation results

[0578] Step 13:

[0579] terminal

[0580] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (graphs, charts, dashboards, etc.). It uses data visualization libraries such as D3.js and Chart.js. It also displays feedback information from the emotion engine.

[0581] Input: Simulation results, emotional feedback

[0582] Data processing: Data visualization, data display

[0583] Output: Visualized simulation results

[0584] (Application Example 2)

[0585] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0586] Conventional emotion recognition technologies and business decision support systems have struggled to reflect users' emotional states in business decisions, resulting in low accuracy in providing feedback for specific business scenarios. Furthermore, the lack of real-time customer service improvements based on customer emotions in physical stores has led to inconsistencies in the quality of the customer experience.

[0587] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0588] This invention includes a server that collects executives' careers, past decisions, statements, and related literature and generates them as a centralized dataset; a server that trains a machine learning model based on the pre-processed data to generate a digital twin of the executives; a server that analyzes data and scenarios related to new management decisions and runs simulations; a server that recognizes the emotions of customers' facial expressions and voices in real time and provides feedback on customer service methods; and a server that visually displays the results of the simulations to the user. This makes it possible to improve the accuracy of management decisions by utilizing the emotional state of the user, and further enables real-time improvement of customer service based on customer emotions in physical stores.

[0589] The term "executive" refers to a person who is in a position to make important decisions regarding the management of a company or organization.

[0590] "Career history" refers to a series of records including an executive's work history, educational history, past duties and responsibilities, and achievements.

[0591] "Past decisions" refers to records of important judgments and decisions made by executives regarding management and operations to date.

[0592] "Records of statements" refer to information that records the opinions and statements made by executives in official settings or meetings.

[0593] "Relevant literature" refers to reference materials such as books, papers, reports, and articles related to executives and management decisions.

[0594] A "centralized dataset" refers to a collection of data compiled from large amounts of information gathered from different sources into a unified format or style.

[0595] A "machine learning model" refers to an algorithm that learns patterns and features based on large amounts of data, and then uses that learning to make predictions and classifications on new data.

[0596] A "digital twin" refers to a digital model that recreates a real-world physical object or person in a virtual space.

[0597] A "scenario" refers to the setting of a simulation that summarizes specific situations and assumptions related to new management decisions or strategies.

[0598] "Simulation" refers to the process of virtually predicting the outcome of business decisions based on a set scenario.

[0599] "Visual display" refers to presenting data and analysis results to users in visual formats such as graphs, charts, and dashboards.

[0600] "Customer" refers to consumers or clients who use the products or services of a company or store.

[0601] "Emotional cues in facial expressions and voice" refers to the inner emotional state that can be gleaned from a customer's facial expressions, tone of voice, and intonation.

[0602] "Customer service methods" refer to the ways in which one responds to and acts to effectively provide services or products to customers.

[0603] "Feedback" refers to the provision of evaluations and advice regarding actions and decisions by systems or individuals.

[0604] "Real-time" refers to the processing and reaction of information occurring at roughly the same time as events in the real world.

[0605] System Overview

[0606] This invention uses a centralized dataset including executives' career histories and past decisions to generate digital twins of executives, simulate new management decisions, and visually display the results to the user. It also includes a system that recognizes the emotions of customers' facial expressions and voices in real time in physical stores and provides feedback on how to serve them.

[0607] Hardware and software configuration

[0608] Hardware:

[0609] Terminal: A digital terminal with advanced display and input devices.

[0610] Camera: High-resolution camera built into the smart glasses.

[0611] Microphone: A high-precision microphone built into the smart glasses.

[0612] software:

[0613] Emotion analysis engine: DeepFace and Google Speech-to-Text for analyzing user and customer facial expressions and voices.

[0614] Natural Language Processing (NLP): The Hugging Face transformers library is used for processing text data.

[0615] Machine learning models: PyTorch and TensorFlow are used for training based on datasets.

[0616] System operation and data processing

[0617] 1. Data Collection: Users input executives' resumes, past decisions, and relevant literature into their devices, and this information is sent to the server as a centralized dataset.

[0618] 2. Digital Twin Generation: Based on the pre-processed data, the server generates digital twins of executives and stores them in the database. This involves the application of NLP techniques such as text normalization, tokenization, stop word removal, and stemming.

[0619] 3. Simulation: Data and scenarios related to new business decisions are entered via a terminal, and the simulation is executed on the server. This generates and visually displays the prediction results.

[0620] 4. Emotion Recognition and Feedback: In physical stores, customer emotions are analyzed in real time from facial expressions and voice, and feedback on customer service methods is displayed on the smart glasses' screen.

[0621] Specific examples

[0622] New product market launch: Users input information about the new product into a terminal, and based on simulations, they can understand predicted sales and risk factors.

[0623] Business acquisition consideration: Users submit information about potential acquisition targets to a server to obtain risk assessments and forecasts of expected returns.

[0624] Improving customer service in physical stores: Employees wearing smart glasses can recognize customer emotions in real time while interacting with customers and receive feedback on how to respond appropriately.

[0625] Example of a prompt

[0626] If a customer says, "This is a bit expensive," and looks dissatisfied:

[0627] "The customer seems dissatisfied. Please suggest other options."

[0628] When simulating the market launch of a new product:

[0629] "Input: Information regarding the market launch of new smart devices. Output: Projected sales, risk factors, and marketing strategy."

[0630] summary

[0631] This invention makes it possible to improve the accuracy of business decisions by utilizing the emotional state of users, and enables real-time improvements to customer service based on customer emotions in physical stores.

[0632] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0633] Step 1:

[0634] Users input executives' career histories, past decisions, statements, and related literature into a terminal to generate a centralized dataset. Input data includes text, scanned documents, etc. The terminal formats this data into a unified format and sends it to the server.

[0635] Step 2:

[0636] The server preprocesses the received data. Specifically, natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming are applied. The preprocessed data is then converted into a format suitable for training machine learning models.

[0637] Step 3:

[0638] The server trains a machine learning model based on pre-processed data to generate a digital twin of the executives. Deep learning and reinforcement learning algorithms are used to train the model. The generated digital twin model learns the executives' behavioral patterns and decision-making processes.

[0639] Step 4:

[0640] The server stores the trained digital twin model in a database, making it available for simulation in new scenarios. The stored model can be quickly accessed for subsequent processing.

[0641] Step 5:

[0642] The user inputs data and scenarios related to new business decisions into the device. This may include scenario information for new product launches or business acquisitions. The user's emotional state is also recognized either through input or automatically by the emotion engine.

[0643] Step 6:

[0644] The terminal sends the newly entered data and scenario, along with the results of the emotion engine's analysis, to the server. The transmitted data forms the basis for the new simulation.

[0645] Step 7:

[0646] The server combines the digital twin model with new data to run simulations. This includes risk assessment, expected returns, and optimal strategies. The results are generated in a format that is visually displayed to the user.

[0647] Step 8:

[0648] The server sends the simulation results to the terminal. The terminal generates graphs, charts, and dashboards to visually display the received results in an easy-to-understand manner.

[0649] Step 9:

[0650] In physical stores, the smart glasses' camera and microphone are used to recognize customers' facial expressions and voices in real time. The emotion engine analyzes the video and audio data to identify the customer's emotions.

[0651] Step 10:

[0652] Based on the emotion analysis results, the server generates appropriate customer service feedback and displays it on the smart glasses' display. If the customer is happy, it will provide specific instructions such as "Please continue to assist them," while if they are dissatisfied, it will provide instructions such as "Please try offering a different suggestion."

[0653] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating 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.

[0654] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0655] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0656] [Second Embodiment]

[0657] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0658] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0659] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0660] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0661] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0662] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0663] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0664] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0665] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0666] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0667] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0668] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. 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".

[0669] This invention is a system for digitizing executives' expertise and supporting management decisions, and a specific embodiment thereof is shown below.

[0670] System Configuration

[0671] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[0672] Program processing

[0673] Data collection and transmission

[0674] User

[0675] Users input data such as executives' resumes, past decisions, statements, and relevant literature into the terminal.

[0676] Users collect information from external data sources and upload it to their devices.

[0677] terminal

[0678] The terminal centrally manages the input data and builds a dataset for transmission to the server.

[0679] The terminal standardizes the data format and corrects inconsistencies.

[0680] server

[0681] The server receives data sent from the terminal and stores it in the database.

[0682] Data preprocessing and model generation

[0683] server

[0684] The server preprocesses the received data using natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[0685] The server trains a machine learning model based on pre-processed data to generate a digital twin of the executives.

[0686] The server saves the generated digital twin model to a database.

[0687] Simulation execution and result display

[0688] User

[0689] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[0690] terminal

[0691] The terminal sends the entered data to the server.

[0692] server

[0693] The server runs simulations using a digital twin model and generates prediction results. These simulation results include risk assessments, expected profits, and optimal strategies.

[0694] The server exports the simulation results as a dataset and sends it to the terminal.

[0695] terminal

[0696] The terminal visually displays the simulation results received from the server. This can be done using formats such as graphs, charts, and dashboards.

[0697] Specific example

[0698] Example 1: Launching a new product into market

[0699] User

[0700] The user enters information about the market launch of a new smart device model into the terminal (e.g., key product features, target market, sales budget).

[0701] terminal

[0702] The terminal normalizes the input data and sends it to the server as a dataset.

[0703] server

[0704] The server runs simulations using digital twin models of executives, generating information such as projected sales, risk factors, and marketing strategies.

[0705] The server sends the simulation results to the terminal.

[0706] terminal

[0707] The terminal visually displays the simulation results received from the server to the user.

[0708] Example 2: Considering a business acquisition

[0709] User

[0710] The user enters information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal.

[0711] terminal

[0712] The terminal sends the input data to the server as a dataset.

[0713] server

[0714] The server runs simulations using digital twin models of executives. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[0715] The server sends the simulation results to the terminal.

[0716] terminal

[0717] The terminal visually displays the simulation results received from the server to the user.

[0718] This system digitizes the knowledge and experience of executives and utilizes it in real time for management decisions, thereby supporting companies in effectively formulating long-term strategies.

[0719] The following describes the processing flow.

[0720] Step 1:

[0721] User

[0722] Users input executives' career histories, past decisions, statements, and relevant documents into the terminal. This includes text data and scanned documents.

[0723] Step 2:

[0724] terminal

[0725] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[0726] Step 3:

[0727] terminal

[0728] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[0729] Step 4:

[0730] terminal

[0731] The terminal sends the centralized dataset to the server. During this process, it detects data transfer errors and attempts to resend the data if necessary.

[0732] Step 5:

[0733] server

[0734] The server receives data sent from the terminal and stores it in a database. This database is used for subsequent processing.

[0735] Step 6:

[0736] server

[0737] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[0738] Step 7:

[0739] server

[0740] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning.

[0741] Step 8:

[0742] server

[0743] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[0744] Step 9:

[0745] server

[0746] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[0747] Step 10:

[0748] User

[0749] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[0750] Step 11:

[0751] terminal

[0752] The terminal sends the new data and scenario entered by the user to the server.

[0753] Step 12:

[0754] server

[0755] The server analyzes new data using a digital twin model and runs simulations. The simulation results include risk assessments, expected profits, and optimal strategies.

[0756] Step 13:

[0757] server

[0758] The server exports the simulation results as a dataset and sends it to the terminal.

[0759] Step 14:

[0760] terminal

[0761] The terminal displays the simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards, etc.).

[0762] Step 15:

[0763] User

[0764] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[0765] (Example 1)

[0766] Next, we will describe Example 1. 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."

[0767] Traditional management decision-making systems struggle to digitize the knowledge and experience of executives and utilize it in real time, making it difficult to improve the quality of management decisions. Furthermore, there is a need for a system that can centrally manage and effectively analyze data on executives' past decisions and statements.

[0768] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0769] This invention includes a server that collects the career history, past decisions, statements, and related literature of executives and generates them as a centralized dataset; a server that normalizes, tokenizes, removes stop words, and stems the data using natural language processing techniques based on the preprocessed data; and a server that trains a machine learning model based on the preprocessed data to generate a digital twin of the executives. This makes it possible to digitize the knowledge and experience of executives and utilize it in management decisions in real time.

[0770] "Executives" refer to individuals who make important decisions within a company or organization, and primarily include management and executives.

[0771] "Career history" refers to the historical information of an employee, including their past job duties, achievements, and educational background.

[0772] "Past decisions" refers to specific decisions, judgments, or statements of intent made by an executive in the past.

[0773] "Record of statements" refers to data in text format that records statements and comments made by officials in the past.

[0774] "Literature" refers to materials and information sources such as books, papers, and business reports related to the person in charge.

[0775] A "dataset" refers to a collection of data that has been centrally managed and organized for use in analysis and model training.

[0776] "Natural language processing technology" refers to techniques for converting text data into a format that is easily understood by machines, and includes normalization, tokenization, stop word removal, stemming, and other methods.

[0777] "Normalization" refers to the process of handling character size, spaces, and special characters in order to unify the format of data.

[0778] "Tokenization" is the process of dividing text into words and phrases, which means breaking down a sentence into units that are easier to analyze.

[0779] "Stop word removal" refers to the process of excluding certain common words (such as "a" or "the") from the analysis.

[0780] "Stemming" is the process of extracting only the stem portion of a word, and it refers to unifying the inflected forms of the word.

[0781] A "machine learning model" refers to an algorithm that learns patterns and regularities based on data and uses that knowledge to make predictions and classifications on new data.

[0782] A "digital twin" is a digital representation that mimics the knowledge and judgment of real-world officials, and is used for simulations and predictions.

[0783] A "scenario" refers to a specific situation or plan related to a new management decision.

[0784] "Simulation" refers to using digital twin models to make predictions and analyses about new situations and scenarios.

[0785] "Visualization" refers to techniques that visually display simulation results using graphs, charts, and other methods to make them easier to understand.

[0786] This invention is a system for digitizing the knowledge of executives and supporting management decisions. Its specific embodiments are as follows:

[0787] System Configuration

[0788] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[0789] Data collection and transmission

[0790] User

[0791] Users input data such as the career history of executives, past decisions, records of statements, and relevant literature into spreadsheets or dedicated forms. They also collect information from external data sources using APIs and scraping tools and upload it to their devices.

[0792] terminal

[0793] The terminal converts the input data into a CSV file and builds a dataset for centralized management. It standardizes the data format using Excel or Google Sheets, detects and corrects inconsistent data, and sends the corrected data to the server.

[0794] server

[0795] The server receives data sent from the terminal and saves it to a dedicated database (e.g., PostgreSQL). During this process, it checks the data's integrity to ensure there is no invalid data.

[0796] Data preprocessing

[0797] server

[0798] The server uses natural language processing techniques (e.g., Python's NLP library spaCy) to normalize, tokenize, remove stop words from, and stem the stored data. For example, all characters are converted to lowercase, unnecessary spaces and special characters are removed, and then tokenization is performed. This divides the data into units of words and phrases, common words (stop words) are removed, and stemming is performed to extract only the word stems.

[0799] Generation of a digital twin model

[0800] server

[0801] The server trains a digital twin model using machine learning algorithms (e.g., TensorFlow, scikit-learn) based on preprocessed data. For example, it uses random forests or neural networks to generate a digital twin, which is a digital representation that mimics the knowledge of the person in charge. The trained model is stored in a dedicated database for performance evaluation and optimization.

[0802] Running the simulation and displaying the results.

[0803] User

[0804] The user inputs data and scenarios related to new business decisions into the terminal. For example, they might input a specific scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[0805] terminal

[0806] The terminal converts the entered scenario data into JSON format and sends it to the server.

[0807] server

[0808] The server uses a digital twin model to perform scenario-based simulations and generates predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising" and send them to the terminal.

[0809] terminal

[0810] The terminal visually displays the received simulation results. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[0811] Specific example

[0812] Example 1: Launching a new product into market

[0813] User

[0814] The user inputs information about the market launch of a new smart device model into the terminal. For example, they might input, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[0815] terminal

[0816] The terminal converts the input data into JSON format and sends it to the server.

[0817] server

[0818] The server runs a simulation using a digital twin model, generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," and sends them to the terminal.

[0819] terminal

[0820] The terminal displays the received results in graphs. Sales forecasts are shown as bar graphs, risk assessments as heatmaps, and recommended strategies as lists.

[0821] Example 2: Considering a business acquisition

[0822] User

[0823] The user inputs the following data into the terminal: "Company Z is being considered for acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[0824] terminal

[0825] The terminal converts this data into a CSV file and sends it to the server.

[0826] server

[0827] The server runs a simulation using a digital twin model, generates results such as "Risk assessment: High, Expected return: 2 billion yen, Recommended strategy: Enhanced due diligence," and sends them to the terminal.

[0828] terminal

[0829] The terminal displays the received results in charts. Risk assessment is shown as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[0830] This will enable the digitization of executives' knowledge and experience, allowing them to be used in real-time management decisions. The system will support companies in effectively formulating long-term strategies.

[0831] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0832] Step 1: Data collection and input

[0833] User

[0834] Users input data such as the career history, past decisions, statements, and relevant literature of executives into spreadsheets or dedicated forms. For example, a user might input "Executive A's career history, past decisions, and relevant literature" into a spreadsheet. They can also collect information from external data sources using APIs or scraping tools and upload it to their device. The input data is saved in CSV or Excel format.

[0835] terminal

[0836] The terminal receives the input data, converts it into a CSV file, and builds a centrally managed dataset. For example, it standardizes the data format using Excel or Google Sheets, detects inconsistent data, and corrects it. The corrected data is prepared as a temporary storage file.

[0837] Input: Records of the person's career history, past decisions, statements, and relevant literature (CSV format file)

[0838] Output: Centrally managed dataset (CSV format file)

[0839] Step 2: Send

[0840] terminal

[0841] The terminal sends a centrally managed dataset to the server. During this process, the integrity of the files is checked, and retransmission is performed if necessary. For example, a data transmission API is used to send the dataset to the server.

[0842] Input: Centrally managed dataset (CSV format file)

[0843] Output: Request to send data to the server

[0844] server

[0845] The server receives data sent from the terminal and stores it in a database. For example, it might store it in a PostgreSQL database and perform data integrity checks.

[0846] Input: Data transmission request

[0847] Output: Data stored in the database

[0848] Step 3: Data Preprocessing

[0849] server

[0850] The server preprocesses the stored data using natural language processing techniques. First, it performs text normalization, converting all characters to lowercase and removing unnecessary spaces and special characters. Next, it performs tokenization, dividing the data into units of words and phrases. Furthermore, it removes stop words and extracts the stem portion of words using stemming.

[0851] Input: Data stored in the database

[0852] Output: Preprocessed text data

[0853] Specific operation: Use the Python spaCy library to perform text normalization, tokenization, stop word removal, and stemming.

[0854] Step 4: Generating a Digital Twin Model

[0855] server

[0856] The server trains a digital twin model using machine learning algorithms based on pre-processed data. Specifically, it uses random forest and neural network algorithms to generate a digital twin that mimics the knowledge of the executives. The trained model undergoes performance evaluation and optimization before being stored in a database.

[0857] Input: Preprocessed text data

[0858] Output: Trained digital twin model

[0859] Specific actions: Train the model using TensorFlow or scikit-learn.

[0860] Step 5: Run the simulation

[0861] User

[0862] The user inputs a scenario related to a new business decision into the terminal. For example, they might input a scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[0863] terminal

[0864] The terminal converts the entered scenario data into JSON format and sends it to the server.

[0865] Input: Scenario (text format)

[0866] Output: Scenario data in JSON format

[0867] server

[0868] The server uses a digital twin model to perform scenario-based simulations and generate predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising."

[0869] Input: Scenario data in JSON format

[0870] Output: Simulation results (JSON format)

[0871] Step 6: Displaying the simulation results

[0872] terminal

[0873] The terminal visually displays the simulation results received from the server. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[0874] Input: Simulation results (JSON format)

[0875] Output: Visualized simulation results (graphs, heatmaps, lists)

[0876] Specific example

[0877] Example 1: Launching a new product into market

[0878] User

[0879] The user enters the following information into the terminal: "New product Y to be launched, target market: men in their 20s, sales budget: 30 million yen."

[0880] terminal

[0881] The device normalizes this information, converts it to JSON format, and sends it to the server.

[0882] server

[0883] The server runs a simulation using a digital twin model and generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," which it then sends to the terminal.

[0884] terminal

[0885] The device visually displays the results, showing sales forecasts as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[0886] Example 2: Considering a business acquisition

[0887] User

[0888] The user enters the following information into the terminal: "Company Z is considering acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[0889] terminal

[0890] The terminal normalizes this information, converts it to CSV format, and sends it to the server.

[0891] server

[0892] The server runs a simulation using a digital twin model and generates results such as "Risk Assessment: High, Expected Return: 2 billion yen, Recommended Strategy: Enhanced Due Diligence," which it then sends to the terminal.

[0893] terminal

[0894] The device visually displays the results, showing risk assessments as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[0895] Following these specific processing steps, the system can digitize the knowledge and experience of executives and utilize it in real time for management decisions.

[0896] (Application Example 1)

[0897] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0898] In modern corporate management, there is a need to leverage the insights of executives and management in real time to effectively support management decisions. Furthermore, it is crucial to efficiently collect operational data from machinery and equipment used in factories and formulate optimal operational plans. However, systems to address these challenges are still not adequately developed, resulting in insufficient improvements in the quality of management decisions and the efficiency of factory operations.

[0899] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0900] This invention includes a server that collects executives' careers, past decisions, records of statements, and related literature and generates them as a centralized dataset; a server that trains a machine learning model based on the pre-processed data to generate a digital twin of the executives; a server that collects operational data of machinery and equipment used in the factory and generates an optimized operational plan; a server that analyzes data and scenarios related to new management decisions and runs simulations; and a server that visually displays the results of the simulations to the user. This enables the real-time utilization of management insights to support management decisions and further improves and optimizes the efficiency of factory operations.

[0901] An "executive" is a person who is in a position to make important decisions within a company or organization and who possesses extensive knowledge and experience in management.

[0902] "Background" refers to information that shows what kind of jobs and roles a person has experienced in the past.

[0903] A "record of statements" is data that compiles and saves statements and comments made by a specific person in the past.

[0904] "Related literature" refers to documentary materials such as books, articles, and reports related to the person or event in question.

[0905] A "centralized dataset" is a collection of data that integrates data gathered from multiple different sources and organizes it into a manageable format.

[0906] A "machine learning model" is a data analysis algorithm created based on collected data to perform predictions and classifications.

[0907] A "digital twin" is a virtual replica of a real-world object, such as a data-based imitation of an executive's knowledge and experience.

[0908] "Machinery and equipment used in factories" refers to the machinery and equipment used in manufacturing and production processes.

[0909] "Operational data" refers to specific numerical data related to the operational status of machinery and equipment, such as operating status, production volume, and the number of errors that occurred.

[0910] An "optimized operational plan" is a plan formulated with the aim of ensuring the efficient and effective operation of machinery and equipment.

[0911] "Data related to new management decisions" refers to data that shows information and scenarios about new business strategies, investment plans, etc.

[0912] A "simulation" is a test or experiment that simulates real-world situations, and is a method for making predictions or evaluations under specific conditions.

[0913] "Means of visual display" refers to methods of displaying data and simulation results in a format that is easy for users to understand, and includes graphs, charts, dashboards, and so on.

[0914] This invention is a system for digitizing executive knowledge and supporting management decisions. The system consists of three components: a terminal, a server, and a user. A specific embodiment of this system is described below.

[0915] System Configuration

[0916] This system primarily uses the following hardware and software:

[0917] Hardware: Smartphones and tablets (devices), servers

[0918] Software: Python, Pandas, Scikit-learn, Matplotlib

[0919] Program processing

[0920] Data collection and transmission

[0921] Users input executives' career histories, past decisions, statements, and relevant literature into a terminal using a smartphone or tablet. Users also input operational data for machinery used in the factory (e.g., operating hours, number of errors, production quantity). The terminal normalizes the input data and sends it to the server as a centralized dataset.

[0922] Data preprocessing and model generation

[0923] The server receives data sent from terminals and performs preprocessing such as text normalization, tokenization, and standardization. The server then trains machine learning models based on the preprocessed data to generate digital twins of executives. Additionally, the server generates optimized operational plans using operational data from machinery used in factories and saves each model.

[0924] Simulation execution and result display

[0925] Users input data and scenarios related to new business decisions into a terminal. Examples include scenarios for new product launches or business acquisitions, annual production targets, and maintenance schedules. The terminal sends this data to a server. The server runs simulations using a digital twin model and operational data model to generate forecast results. These forecast results include risk assessments, expected profits, and production plans. The server sends the simulation results back to the terminal, which displays them visually. Display formats include graphs, charts, and dashboards.

[0926] Specific example: Optimizing the annual operational plan for factory robots.

[0927] The user inputs annual operational data (e.g., monthly production targets, maintenance schedules for various machines, etc.). The following prompts are used as an example:

[0928] Example of an input prompt:

[0929] Please enter the following operational data:

[0930] Months: January, February, March...December

[0931] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[0932] Maintenance scheduled for: January 15th, February 20th...

[0933] Starting the simulation.

[0934] This allows users to input necessary data through their terminals, and the server to run simulations using operational data models, generating and presenting optimal operational plans.

[0935] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0936] Step 1:

[0937] Users input executives' career histories, past decisions, statements, relevant literature, and operational data of machinery used in the factory into a terminal. The entered data includes executive attribute information (e.g., name, past duties), details of statements, literature citations, working hours, number of errors, and production quantities. The terminal centrally collects this data, normalizes it as needed, and corrects inconsistencies.

[0938] Step 2:

[0939] The device collects and normalizes the data, constructs it as a dataset, and sends it to the server. During this process, the data is converted to a standardized format such as CSV or JSON. The receiving server receives this data and stores it in its database.

[0940] Step 3:

[0941] The server preprocesses the received data. This includes natural language processing (NLP) techniques such as text normalization, tokenization, and stop word removal. Numerical data is also preprocessed through standardization and normalization. The preprocessed data is converted into a parseable format and proceeds to the next processing step.

[0942] Step 4:

[0943] The server trains machine learning models using pre-processed data. These models include digital twin models of executives and models that generate optimal operational plans for machinery and equipment used in factories. These models are trained using random forest regression and deep learning techniques. The generated models are stored on the server.

[0944] Step 5:

[0945] The user inputs data and scenarios related to new business decisions into the terminal. Specifically, they provide scenario information such as new product launches, business acquisitions, and new production targets. The following prompts are used:

[0946] Example of an input prompt:

[0947] Please enter the following operational data:

[0948] Months: January, February, March...December

[0949] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[0950] Maintenance scheduled for: January 15th, February 20th...

[0951] Starting the simulation.

[0952] Step 6:

[0953] The terminal sends the new data and scenario entered by the user to the server. The server receives this data, performs preprocessing and standardization, and then runs a simulation using the stored digital twin model and operational planning model.

[0954] Step 7:

[0955] The server generates the results of the executed simulation. These predictions include risk assessments, expected profits, and optimized production plans. The simulation results are stored as a dataset on the server and sent to the terminal.

[0956] Step 8:

[0957] The terminal visually displays simulation results received from the server to the user. This includes graphs, charts, and dashboards. Users can use this information to make specific business decisions and operational plan decisions.

[0958] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0959] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions, and a specific embodiment thereof is shown below.

[0960] System Configuration

[0961] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[0962] Program processing

[0963] Data collection and transmission

[0964] User

[0965] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Furthermore, users input their own emotional state or have it automatically recognized by an emotion engine.

[0966] terminal

[0967] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted for subsequent processing. The device also sends emotion recognition requests to the emotion engine.

[0968] Emotional Engine

[0969] The emotion engine analyzes the user's emotions using user input data or facial and speech recognition technologies. The analysis results are sent to a server for later use in simulations.

[0970] server

[0971] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database.

[0972] Data preprocessing and model generation

[0973] server

[0974] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[0975] server

[0976] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[0977] server

[0978] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[0979] server

[0980] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[0981] Simulation execution and result display

[0982] User

[0983] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[0984] terminal

[0985] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[0986] server

[0987] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[0988] server

[0989] The server exports the simulation results as a dataset and sends it to the terminal.

[0990] terminal

[0991] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[0992] Specific example

[0993] Example 1: Launching a new product into market

[0994] User

[0995] The user inputs information about the market launch of the new smart device (e.g., key product features, target market, sales budget) into the device. Emotional states are also input simultaneously or recognized by the emotion engine.

[0996] terminal

[0997] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[0998] server

[0999] The server runs simulations using digital twin models of executives and data from the emotion engine. It generates information such as predicted sales, risk factors, and marketing strategies.

[1000] server

[1001] The server sends the simulation results to the terminal.

[1002] terminal

[1003] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1004] Example 2: Considering a business acquisition

[1005] User

[1006] The user inputs information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal. Emotional states are also input simultaneously or recognized by an emotion engine.

[1007] terminal

[1008] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[1009] server

[1010] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[1011] server

[1012] The server sends the simulation results to the terminal.

[1013] terminal

[1014] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1015] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[1016] The following describes the processing flow.

[1017] Step 1:

[1018] User

[1019] Users input executives' career histories, past decisions, statements, and relevant literature into the terminal. If emotional states are also to be entered, they can be done using text or a dedicated emotion input interface.

[1020] Step 2:

[1021] terminal

[1022] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[1023] Step 3:

[1024] terminal

[1025] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[1026] Step 4:

[1027] terminal

[1028] The terminal sends the input data to the server. At the same time, it sends an emotion recognition request to the emotion engine.

[1029] Step 5:

[1030] Emotional Engine

[1031] The emotion engine analyzes the user's emotions using user input data, facial recognition, and speech recognition technology. The analysis results are then sent to the server.

[1032] Step 6:

[1033] server

[1034] The server receives data sent from the terminal and the emotion engine and stores it in a database. This database is used for subsequent processing.

[1035] Step 7:

[1036] server

[1037] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[1038] Step 8:

[1039] server

[1040] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[1041] Step 9:

[1042] server

[1043] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[1044] Step 10:

[1045] server

[1046] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[1047] Step 11:

[1048] User

[1049] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[1050] Step 12:

[1051] terminal

[1052] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[1053] Step 13:

[1054] server

[1055] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[1056] Step 14:

[1057] server

[1058] The server exports the simulation results as a dataset and sends it to the terminal.

[1059] Step 15:

[1060] terminal

[1061] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[1062] Step 16:

[1063] User

[1064] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[1065] (Example 2)

[1066] Next, we will describe Example 2. 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".

[1067] In recent years, it has become increasingly important for companies to consider not only the experience and knowledge of executives but also the emotions and subjectivity of users when making management decisions. However, there is a lack of systems that can properly centralize, analyze, and visually display the results of simulations based on this information. As a result, the management decision-making process is not conducted efficiently and effectively, leading to a decline in the quality of decision-making in corporate operations.

[1068] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1069] In this invention, the server includes means for inputting or recognizing the user's emotional state, means for collecting the executive's career history, past decisions, records of statements, and related literature, and for formatting and storing them as a centralized dataset, means for training a machine learning model based on the pre-processed data and generating a digital twin of the executive, means for analyzing data and scenarios related to new management decisions while considering the user's emotional state, and for running simulations, and means for exporting the simulation results as a dataset and further displaying it visually. This makes it possible to comprehensively manage and analyze diverse information and support management decisions that reflect emotions.

[1070] "User emotional state" refers to the user's emotions and psychological state, and includes information input through the human-machine interface and emotional information recognized using an emotion engine.

[1071] A "centralized dataset" is a collection of data that integrates and manages executives' career histories, past decisions, statements, and related literature, and then formats and stores them in a format suitable for analysis and processing.

[1072] "Preprocessed data" refers to data that has been transformed into a format suitable for training machine learning models by applying processing such as normalization, tokenization, stop word removal, and stemming to the input data using techniques such as natural language processing.

[1073] A "machine learning model" is an algorithm and system that learns patterns from data and uses that knowledge to make predictions, classifications, and identifications.

[1074] A "digital twin" is a digital twin model created based on machine learning models to reproduce physical executives, their actions, and decision-making processes in a virtual environment.

[1075] A "scenario" refers to a specific combination of circumstances or conditions that should be considered for business decisions, and specifically describes case studies such as the launch of a new product or the acquisition of a business.

[1076] A "simulation" is a computational process that involves trial and error in a virtual environment before actually performing an action, and uses the results to derive risk assessments and optimal strategies.

[1077] "Visual display methods" refer to ways of providing data and analysis results to users in an intuitive and easy-to-understand manner using formats such as graphs, charts, and dashboards.

[1078] A "generative AI model" is an artificial intelligence model that learns patterns and characteristics of data and makes predictions and suggestions based on new data.

[1079] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions. The system is configured as follows:

[1080] System Configuration

[1081] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[1082] Program processing

[1083] Data collection and transmission

[1084] User

[1085] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Users also input their own emotional state or undergo automatic recognition by an emotion engine.

[1086] terminal

[1087] The terminal centrally manages user input data and builds it into a dataset. The data is formatted into an appropriate format. Furthermore, the terminal sends emotion recognition requests to the emotion engine. A typical PC or mobile device can be used as the terminal.

[1088] Emotional Engine

[1089] The emotion engine analyzes user emotions using user input data, facial recognition, and speech recognition technologies. Specifically, it performs facial recognition using Python and the OpenCV library, and speech analysis using Praat. The analysis results are sent to the server in real time.

[1090] server

[1091] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL).

[1092] Data preprocessing and model generation

[1093] server

[1094] The server preprocesses the received data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming. The preprocessed data is then used to train a machine learning model using TensorFlow or PyTorch. Furthermore, the user's emotional state is also utilized as a parameter in the model.

[1095] server

[1096] The server optimizes the model parameters using a dataset to generate a digital twin for each executive. During this process, the executives' behavioral patterns and decision-making processes are learned. The optimized digital twin model is then stored in a database.

[1097] Simulation execution and result display

[1098] User

[1099] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[1100] terminal

[1101] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[1102] server

[1103] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[1104] server

[1105] The server exports the simulation results as a dataset and sends those results to the terminal.

[1106] terminal

[1107] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (graphs, charts, dashboards, etc.). Data visualization libraries such as D3.js and Chart.js are used for display. In addition, feedback information from the emotion engine is also displayed.

[1108] Specific example

[1109] Example 1: Launching a new product into market

[1110] User

[1111] Users input information about the market launch of a new smart device (key product features, target market, sales budget) into the device. Their emotional state is also input simultaneously or recognized by an emotion engine.

[1112] terminal

[1113] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[1114] server

[1115] The server runs simulations using executive digital twin models and emotion engine data, generating information on predicted sales, risk factors, and marketing strategies.

[1116] server

[1117] The server sends the simulation results to the terminal.

[1118] terminal

[1119] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1120] Example 2: Considering a business acquisition

[1121] User

[1122] The user inputs information about the target company (financial status, synergy effects, acquisition price) into the terminal. Their emotional state is also entered simultaneously or recognized by an emotion engine.

[1123] terminal

[1124] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[1125] server

[1126] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results for acquisition risk assessment, expected returns, and optimal strategies.

[1127] server

[1128] The server sends the simulation results to the terminal.

[1129] terminal

[1130] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1131] Example of a prompt

[1132] 1. "A user is considering launching a new product. Use an emotion engine to analyze the user's emotional state and display simulation results to support their decision-making regarding market launch."

[1133] 2. "A user is considering acquiring a certain company. Based on the target company's data and the results of the sentiment engine, perform a simulation of acquisition risk and expected return, and propose the optimal strategy."

[1134] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[1135] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1136] The processing flow of a system program is divided into processing steps.

[1137] Step 1:

[1138] User

[1139] Users input data such as executives' career histories, past decisions, statements, and relevant literature into the terminal. This input includes text data and scanned documents. In addition, users can manually input their own emotional state or have it automatically recognized by an emotion engine.

[1140] Input: Executive data, literature data, emotional state

[1141] Output: Centralized dataset

[1142] Step 2:

[1143] terminal

[1144] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted into an appropriate format. The device temporarily stores the formatted data and sends an emotion recognition request to the emotion engine.

[1145] Input: User data, emotional state

[1146] Data processing: Data format conversion, data formatting.

[1147] Output: Formatted dataset, sentiment recognition requests

[1148] Step 3:

[1149] Emotional Engine

[1150] The emotion engine analyzes user input data and video and audio data obtained from the device's camera and microphone. It uses facial recognition technology and voice analysis (voice feature extraction and emotion classification) to quantify the user's emotional state. The analysis results are transmitted to the server in real time.

[1151] Input: Video data, audio data

[1152] Data processing: facial recognition, voice analysis

[1153] Output: Sentiment analysis results

[1154] Step 4:

[1155] server

[1156] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL).

[1157] Input: Formatted dataset, sentiment analysis results

[1158] Data processing: Data reception, data storage

[1159] Output: Saved database

[1160] Step 5:

[1161] server

[1162] The server preprocesses the stored data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming.

[1163] Input: Saved data

[1164] Data processing: Text preprocessing (normalization, tokenization, stop word removal, stemming)

[1165] Output: Preprocessed data

[1166] Step 6:

[1167] server

[1168] The server trains machine learning models using TensorFlow or PyTorch based on preprocessed data. Furthermore, it utilizes the user's emotional state as a parameter in the model.

[1169] Input: Preprocessed data, emotional state

[1170] Data processing: Machine learning model training

[1171] Output: Trained machine learning model

[1172] Step 7:

[1173] server

[1174] The server optimizes the parameters of the digital twin model using executive data. In this process, the executives' behavioral patterns and decision-making processes are learned.

[1175] Input: Executive data, machine learning model

[1176] Data Calculation: Model Parameter Optimization

[1177] Output: Optimized digital twin model

[1178] Step 8:

[1179] server

[1180] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[1181] Input: Optimized digital twin model

[1182] Data processing: Data storage

[1183] Output: Saved digital twin model

[1184] Step 9:

[1185] User

[1186] Users input data and scenarios related to new business decisions into the terminal. Specific examples include information on new product launches and business acquisitions. Emotional states are either entered sequentially or automatically updated by the emotion engine.

[1187] Input: New scenario data, emotional state

[1188] Output: Input scenario data

[1189] Step 10:

[1190] terminal

[1191] The device sends new scenario data and sentiment analysis results to the server.

[1192] Input: Input scenario data, sentiment analysis results

[1193] Data processing: Data transmission

[1194] Output: Sent data

[1195] Step 11:

[1196] server

[1197] The server uses information from the digital twin model and the emotion engine to perform simulations and generate prediction results. Specific examples include risk assessment, expected profit, and optimal strategy.

[1198] Input: Digital twin model, scenario data, sentiment analysis results

[1199] Data calculation: Simulation execution

[1200] Output: Simulation results

[1201] Step 12:

[1202] server

[1203] The server exports the simulation results as a dataset and sends it to the terminal.

[1204] Input: Simulation results

[1205] Data processing: Data export, data transmission

[1206] Output: Sent simulation results

[1207] Step 13:

[1208] terminal

[1209] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (graphs, charts, dashboards, etc.). It uses data visualization libraries such as D3.js and Chart.js. It also displays feedback information from the emotion engine.

[1210] Input: Simulation results, emotional feedback

[1211] Data processing: Data visualization, data display

[1212] Output: Visualized simulation results

[1213] (Application Example 2)

[1214] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[1215] Conventional emotion recognition technologies and business decision support systems have struggled to reflect users' emotional states in business decisions, resulting in low accuracy in providing feedback for specific business scenarios. Furthermore, the lack of real-time customer service improvements based on customer emotions in physical stores has led to inconsistencies in the quality of the customer experience.

[1216] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1217] This invention includes a server that collects executives' careers, past decisions, statements, and related literature and generates them as a centralized dataset; a server that trains a machine learning model based on the pre-processed data to generate a digital twin of the executives; a server that analyzes data and scenarios related to new management decisions and runs simulations; a server that recognizes the emotions of customers' facial expressions and voices in real time and provides feedback on customer service methods; and a server that visually displays the results of the simulations to the user. This makes it possible to improve the accuracy of management decisions by utilizing the emotional state of the user, and further enables real-time improvement of customer service based on customer emotions in physical stores.

[1218] The term "executive" refers to a person who is in a position to make important decisions regarding the management of a company or organization.

[1219] "Career history" refers to a series of records including an executive's work history, educational history, past duties and responsibilities, and achievements.

[1220] "Past decisions" refers to records of important judgments and decisions made by executives regarding management and operations to date.

[1221] "Records of statements" refer to information that records the opinions and statements made by executives in official settings or meetings.

[1222] "Relevant literature" refers to reference materials such as books, papers, reports, and articles related to executives and management decisions.

[1223] A "centralized dataset" refers to a collection of data compiled from large amounts of information gathered from different sources into a unified format or style.

[1224] A "machine learning model" refers to an algorithm that learns patterns and features based on large amounts of data, and then uses that learning to make predictions and classifications on new data.

[1225] A "digital twin" refers to a digital model that recreates a real-world physical object or person in a virtual space.

[1226] A "scenario" refers to the setting of a simulation that summarizes specific situations and assumptions related to new management decisions or strategies.

[1227] "Simulation" refers to the process of virtually predicting the outcome of business decisions based on a set scenario.

[1228] "Visual display" refers to presenting data and analysis results to users in visual formats such as graphs, charts, and dashboards.

[1229] "Customer" refers to consumers or clients who use the products or services of a company or store.

[1230] "Emotional cues in facial expressions and voice" refers to the inner emotional state that can be gleaned from a customer's facial expressions, tone of voice, and intonation.

[1231] "Customer service methods" refer to the ways in which one responds to and acts to effectively provide services or products to customers.

[1232] "Feedback" refers to the provision of evaluations and advice regarding actions and decisions by systems or individuals.

[1233] "Real-time" refers to the processing and reaction of information occurring at roughly the same time as events in the real world.

[1234] System Overview

[1235] This invention uses a centralized dataset including executives' career histories and past decisions to generate digital twins of executives, simulate new management decisions, and visually display the results to the user. It also includes a system that recognizes the emotions of customers' facial expressions and voices in real time in physical stores and provides feedback on how to serve them.

[1236] Hardware and software configuration

[1237] Hardware:

[1238] Terminal: A digital terminal with advanced display and input devices.

[1239] Camera: High-resolution camera built into the smart glasses.

[1240] Microphone: A high-precision microphone built into the smart glasses.

[1241] software:

[1242] Emotion analysis engine: DeepFace and Google Speech-to-Text for analyzing user and customer facial expressions and voices.

[1243] Natural Language Processing (NLP): The Hugging Face transformers library is used for processing text data.

[1244] Machine learning models: PyTorch and TensorFlow are used for training based on datasets.

[1245] System operation and data processing

[1246] 1. Data Collection: Users input executives' resumes, past decisions, and relevant literature into their devices, and this information is sent to the server as a centralized dataset.

[1247] 2. Digital Twin Generation: Based on the pre-processed data, the server generates digital twins of the executives and stores them in the database. This involves the application of NLP techniques such as text normalization, tokenization, stop word removal, and stemming.

[1248] 3. Simulation: Data and scenarios related to new business decisions are entered via a terminal, and the simulation is executed on the server. This generates and visually displays the prediction results.

[1249] 4. Emotion Recognition and Feedback: In physical stores, customer emotions are analyzed in real time from facial expressions and voice, and feedback on customer service methods is displayed on the smart glasses' screen.

[1250] Specific examples

[1251] New product market launch: Users input information about the new product into a terminal, and based on simulations, they can understand predicted sales and risk factors.

[1252] Business acquisition consideration: Users submit information about potential acquisition targets to a server to obtain risk assessments and forecasts of expected returns.

[1253] Improving customer service in physical stores: Employees wearing smart glasses can recognize customer emotions in real time while interacting with customers and receive feedback on how to respond appropriately.

[1254] Example of a prompt

[1255] If a customer says, "This is a bit expensive," and looks dissatisfied:

[1256] "The customer seems dissatisfied. Please suggest other options."

[1257] When simulating the market launch of a new product:

[1258] "Input: Information regarding the market launch of new smart devices. Output: Projected sales, risk factors, and marketing strategy."

[1259] summary

[1260] This invention makes it possible to improve the accuracy of business decisions by utilizing the emotional state of users, and enables real-time improvements to customer service based on customer emotions in physical stores.

[1261] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1262] Step 1:

[1263] Users input executives' career histories, past decisions, statements, and related literature into a terminal to generate a centralized dataset. Input data includes text, scanned documents, etc. The terminal formats this data into a unified format and sends it to the server.

[1264] Step 2:

[1265] The server preprocesses the received data. Specifically, natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming are applied. The preprocessed data is then converted into a format suitable for training machine learning models.

[1266] Step 3:

[1267] The server trains a machine learning model based on pre-processed data to generate a digital twin of the executives. Deep learning and reinforcement learning algorithms are used to train the model. The generated digital twin model learns the executives' behavioral patterns and decision-making processes.

[1268] Step 4:

[1269] The server stores the trained digital twin model in a database, making it available for simulation in new scenarios. The stored model can be quickly accessed for subsequent processing.

[1270] Step 5:

[1271] The user inputs data and scenarios related to new business decisions into the device. This may include scenario information for new product launches or business acquisitions. The user's emotional state is also recognized either through input or automatically by the emotion engine.

[1272] Step 6:

[1273] The terminal sends the newly entered data and scenario, along with the results of the emotion engine's analysis, to the server. The transmitted data forms the basis for the new simulation.

[1274] Step 7:

[1275] The server combines the digital twin model with new data to run simulations. This includes risk assessment, expected returns, and optimal strategies. The results are generated in a format that is visually displayed to the user.

[1276] Step 8:

[1277] The server sends the simulation results to the terminal. The terminal generates graphs, charts, and dashboards to visually display the received results in an easy-to-understand manner.

[1278] Step 9:

[1279] In physical stores, the smart glasses' camera and microphone are used to recognize customers' facial expressions and voices in real time. The emotion engine analyzes the video and audio data to identify the customer's emotions.

[1280] Step 10:

[1281] Based on the emotion analysis results, the server generates appropriate customer service feedback and displays it on the smart glasses' display. If the customer is happy, it will provide specific instructions such as "Please continue to assist them," while if they are dissatisfied, it will provide instructions such as "Please try offering a different suggestion."

[1282] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1283] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1284] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[1285] [Third Embodiment]

[1286] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[1287] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1288] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1289] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[1290] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1291] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1292] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1293] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1294] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[1295] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1296] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1297] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[1298] This invention is a system for digitizing executives' expertise and supporting management decisions, and a specific embodiment thereof is shown below.

[1299] System Configuration

[1300] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[1301] Program processing

[1302] Data collection and transmission

[1303] User

[1304] Users input data such as executives' resumes, past decisions, statements, and relevant literature into the terminal.

[1305] Users collect information from external data sources and upload it to their devices.

[1306] terminal

[1307] The terminal centrally manages the input data and builds a dataset for transmission to the server.

[1308] The terminal standardizes the data format and corrects inconsistencies.

[1309] server

[1310] The server receives data sent from the terminal and stores it in the database.

[1311] Data preprocessing and model generation

[1312] server

[1313] The server preprocesses the received data using natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[1314] The server trains a machine learning model based on pre-processed data to generate a digital twin of the executives.

[1315] The server saves the generated digital twin model to a database.

[1316] Simulation execution and result display

[1317] User

[1318] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[1319] terminal

[1320] The terminal sends the entered data to the server.

[1321] server

[1322] The server runs simulations using a digital twin model and generates prediction results. These simulation results include risk assessments, expected profits, and optimal strategies.

[1323] The server exports the simulation results as a dataset and sends it to the terminal.

[1324] terminal

[1325] The terminal visually displays the simulation results received from the server. This can be done using formats such as graphs, charts, and dashboards.

[1326] Specific example

[1327] Example 1: Launching a new product into market

[1328] User

[1329] The user enters information about the market launch of a new smart device model into the terminal (e.g., key product features, target market, sales budget).

[1330] terminal

[1331] The terminal normalizes the input data and sends it to the server as a dataset.

[1332] server

[1333] The server runs simulations using digital twin models of executives, generating information such as projected sales, risk factors, and marketing strategies.

[1334] The server sends the simulation results to the terminal.

[1335] terminal

[1336] The terminal visually displays the simulation results received from the server to the user.

[1337] Example 2: Considering a business acquisition

[1338] User

[1339] The user enters information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal.

[1340] terminal

[1341] The terminal sends the input data to the server as a dataset.

[1342] server

[1343] The server runs simulations using digital twin models of executives. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[1344] The server sends the simulation results to the terminal.

[1345] terminal

[1346] The terminal visually displays the simulation results received from the server to the user.

[1347] This system digitizes the knowledge and experience of executives and utilizes it in real time for management decisions, thereby supporting companies in effectively formulating long-term strategies.

[1348] The following describes the processing flow.

[1349] Step 1:

[1350] User

[1351] Users input executives' career histories, past decisions, statements, and relevant documents into the terminal. This includes text data and scanned documents.

[1352] Step 2:

[1353] terminal

[1354] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[1355] Step 3:

[1356] terminal

[1357] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[1358] Step 4:

[1359] terminal

[1360] The terminal sends the centralized dataset to the server. During this process, it detects data transfer errors and attempts to resend the data if necessary.

[1361] Step 5:

[1362] server

[1363] The server receives data sent from the terminal and stores it in a database. This database is used for subsequent processing.

[1364] Step 6:

[1365] server

[1366] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[1367] Step 7:

[1368] server

[1369] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning.

[1370] Step 8:

[1371] server

[1372] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[1373] Step 9:

[1374] server

[1375] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[1376] Step 10:

[1377] User

[1378] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[1379] Step 11:

[1380] terminal

[1381] The terminal sends the new data and scenario entered by the user to the server.

[1382] Step 12:

[1383] server

[1384] The server analyzes new data using a digital twin model and runs simulations. The simulation results include risk assessments, expected profits, and optimal strategies.

[1385] Step 13:

[1386] server

[1387] The server exports the simulation results as a dataset and sends it to the terminal.

[1388] Step 14:

[1389] terminal

[1390] The terminal displays the simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards, etc.).

[1391] Step 15:

[1392] User

[1393] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[1394] (Example 1)

[1395] Next, we will describe Example 1. 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."

[1396] Traditional management decision-making systems struggle to digitize the knowledge and experience of executives and utilize it in real time, making it difficult to improve the quality of management decisions. Furthermore, there is a need for a system that can centrally manage and effectively analyze data on executives' past decisions and statements.

[1397] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1398] This invention includes a server that collects the career history, past decisions, statements, and related literature of executives and generates them as a centralized dataset; a server that normalizes, tokenizes, removes stop words, and stems the data using natural language processing techniques based on the preprocessed data; and a server that trains a machine learning model based on the preprocessed data to generate a digital twin of the executives. This makes it possible to digitize the knowledge and experience of executives and utilize it in management decisions in real time.

[1399] "Executives" refer to individuals who make important decisions within a company or organization, and primarily include management and executives.

[1400] "Career history" refers to the historical information of an employee, including their past job duties, achievements, and educational background.

[1401] "Past decisions" refers to specific decisions, judgments, or statements of intent made by an executive in the past.

[1402] "Record of statements" refers to data in text format that records statements and comments made by officials in the past.

[1403] "Literature" refers to materials and information sources such as books, papers, and business reports related to the person in charge.

[1404] A "dataset" refers to a collection of data that has been centrally managed and organized for use in analysis and model training.

[1405] "Natural language processing technology" refers to techniques for converting text data into a format that is easily understood by machines, and includes normalization, tokenization, stop word removal, stemming, and other methods.

[1406] "Normalization" refers to the process of handling character size, spaces, and special characters in order to unify the format of data.

[1407] "Tokenization" is the process of dividing text into words and phrases, which means breaking down a sentence into units that are easier to analyze.

[1408] "Stop word removal" refers to the process of excluding certain common words (such as "a" or "the") from the analysis.

[1409] "Stemming" is the process of extracting only the stem portion of a word, and it refers to unifying the inflected forms of the word.

[1410] A "machine learning model" refers to an algorithm that learns patterns and regularities based on data and uses that knowledge to make predictions and classifications on new data.

[1411] A "digital twin" is a digital representation that mimics the knowledge and judgment of real-world officials, and is used for simulations and predictions.

[1412] A "scenario" refers to a specific situation or plan related to a new management decision.

[1413] "Simulation" refers to using digital twin models to make predictions and analyses about new situations and scenarios.

[1414] "Visualization" refers to techniques that visually display simulation results using graphs, charts, and other methods to make them easier to understand.

[1415] This invention is a system for digitizing the knowledge of executives and supporting management decisions. Its specific embodiments are as follows:

[1416] System Configuration

[1417] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[1418] Data collection and transmission

[1419] User

[1420] Users input data such as the career history of executives, past decisions, records of statements, and relevant literature into spreadsheets or dedicated forms. They also collect information from external data sources using APIs and scraping tools and upload it to their devices.

[1421] terminal

[1422] The terminal converts the input data into a CSV file and builds a dataset for centralized management. It standardizes the data format using Excel or Google Sheets, detects and corrects inconsistent data, and sends the corrected data to the server.

[1423] server

[1424] The server receives data sent from the terminal and saves it to a dedicated database (e.g., PostgreSQL). During this process, it checks the data's integrity to ensure there is no invalid data.

[1425] Data preprocessing

[1426] server

[1427] The server uses natural language processing techniques (e.g., Python's NLP library spaCy) to normalize, tokenize, remove stop words from, and stem the stored data. For example, all characters are converted to lowercase, unnecessary spaces and special characters are removed, and then tokenization is performed. This divides the data into units of words and phrases, common words (stop words) are removed, and stemming is performed to extract only the word stems.

[1428] Generation of a digital twin model

[1429] server

[1430] The server trains a digital twin model using machine learning algorithms (e.g., TensorFlow, scikit-learn) based on preprocessed data. For example, it uses random forests or neural networks to generate a digital twin, which is a digital representation that mimics the knowledge of the person in charge. The trained model is stored in a dedicated database for performance evaluation and optimization.

[1431] Running the simulation and displaying the results.

[1432] User

[1433] The user inputs data and scenarios related to new business decisions into the terminal. For example, they might input a specific scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[1434] terminal

[1435] The terminal converts the entered scenario data into JSON format and sends it to the server.

[1436] server

[1437] The server uses a digital twin model to perform scenario-based simulations and generates predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising" and send them to the terminal.

[1438] terminal

[1439] The terminal visually displays the received simulation results. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[1440] Specific example

[1441] Example 1: Launching a new product into market

[1442] User

[1443] The user inputs information about the market launch of a new smart device model into the terminal. For example, they might input, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[1444] terminal

[1445] The terminal converts the input data into JSON format and sends it to the server.

[1446] server

[1447] The server runs a simulation using a digital twin model, generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," and sends them to the terminal.

[1448] terminal

[1449] The terminal displays the received results in graphs. Sales forecasts are shown as bar graphs, risk assessments as heatmaps, and recommended strategies as lists.

[1450] Example 2: Considering a business acquisition

[1451] User

[1452] The user inputs the following data into the terminal: "Company Z is being considered for acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[1453] terminal

[1454] The terminal converts this data into a CSV file and sends it to the server.

[1455] server

[1456] The server runs a simulation using a digital twin model, generates results such as "Risk assessment: High, Expected return: 2 billion yen, Recommended strategy: Enhanced due diligence," and sends them to the terminal.

[1457] terminal

[1458] The terminal displays the received results in charts. Risk assessment is shown as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[1459] This will enable the digitization of executives' knowledge and experience, allowing them to be used in real-time management decisions. The system will support companies in effectively formulating long-term strategies.

[1460] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1461] Step 1: Data collection and input

[1462] User

[1463] Users input data such as the career history, past decisions, statements, and relevant literature of executives into spreadsheets or dedicated forms. For example, a user might input "Executive A's career history, past decisions, and relevant literature" into a spreadsheet. They can also collect information from external data sources using APIs or scraping tools and upload it to their device. The input data is saved in CSV or Excel format.

[1464] terminal

[1465] The terminal receives the input data, converts it into a CSV file, and builds a centrally managed dataset. For example, it standardizes the data format using Excel or Google Sheets, detects inconsistent data, and corrects it. The corrected data is prepared as a temporary storage file.

[1466] Input: Records of the person's career history, past decisions, statements, and relevant literature (CSV format file)

[1467] Output: Centrally managed dataset (CSV format file)

[1468] Step 2: Send

[1469] terminal

[1470] The terminal sends a centrally managed dataset to the server. During this process, the integrity of the files is checked, and retransmission is performed if necessary. For example, a data transmission API is used to send the dataset to the server.

[1471] Input: Centrally managed dataset (CSV format file)

[1472] Output: Request to send data to the server

[1473] server

[1474] The server receives data sent from the terminal and stores it in a database. For example, it might store it in a PostgreSQL database and perform data integrity checks.

[1475] Input: Data transmission request

[1476] Output: Data stored in the database

[1477] Step 3: Data Preprocessing

[1478] server

[1479] The server preprocesses the stored data using natural language processing techniques. First, it performs text normalization, converting all characters to lowercase and removing unnecessary spaces and special characters. Next, it performs tokenization, dividing the data into units of words and phrases. Furthermore, it removes stop words and extracts the stem portion of words using stemming.

[1480] Input: Data stored in the database

[1481] Output: Preprocessed text data

[1482] Specific operation: Use the Python spaCy library to perform text normalization, tokenization, stop word removal, and stemming.

[1483] Step 4: Generating a Digital Twin Model

[1484] server

[1485] The server trains a digital twin model using machine learning algorithms based on pre-processed data. Specifically, it uses random forest and neural network algorithms to generate a digital twin that mimics the knowledge of the executives. The trained model undergoes performance evaluation and optimization before being stored in a database.

[1486] Input: Preprocessed text data

[1487] Output: Trained digital twin model

[1488] Specific actions: Train the model using TensorFlow or scikit-learn.

[1489] Step 5: Run the simulation

[1490] User

[1491] The user inputs a scenario related to a new business decision into the terminal. For example, they might input a scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[1492] terminal

[1493] The terminal converts the entered scenario data into JSON format and sends it to the server.

[1494] Input: Scenario (text format)

[1495] Output: Scenario data in JSON format

[1496] server

[1497] The server uses a digital twin model to perform scenario-based simulations and generate predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising."

[1498] Input: Scenario data in JSON format

[1499] Output: Simulation results (JSON format)

[1500] Step 6: Displaying the simulation results

[1501] terminal

[1502] The terminal visually displays the simulation results received from the server. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[1503] Input: Simulation results (JSON format)

[1504] Output: Visualized simulation results (graphs, heatmaps, lists)

[1505] Specific example

[1506] Example 1: Launching a new product into market

[1507] User

[1508] The user enters the following information into the terminal: "New product Y to be launched, target market: men in their 20s, sales budget: 30 million yen."

[1509] terminal

[1510] The device normalizes this information, converts it to JSON format, and sends it to the server.

[1511] server

[1512] The server runs a simulation using a digital twin model and generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," which it then sends to the terminal.

[1513] terminal

[1514] The device visually displays the results, showing sales forecasts as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[1515] Example 2: Considering a business acquisition

[1516] User

[1517] The user enters the following information into the terminal: "Company Z is considering acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[1518] terminal

[1519] The terminal normalizes this information, converts it to CSV format, and sends it to the server.

[1520] server

[1521] The server runs a simulation using a digital twin model and generates results such as "Risk Assessment: High, Expected Return: 2 billion yen, Recommended Strategy: Enhanced Due Diligence," which it then sends to the terminal.

[1522] terminal

[1523] The device visually displays the results, showing risk assessments as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[1524] Following these specific processing steps, the system can digitize the knowledge and experience of executives and utilize it in real time for management decisions.

[1525] (Application Example 1)

[1526] Next, we will explain Application Example 1. In the following explanation, 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."

[1527] In modern corporate management, there is a need to leverage the insights of executives and management in real time to effectively support management decisions. Furthermore, it is crucial to efficiently collect operational data from machinery and equipment used in factories and formulate optimal operational plans. However, systems to address these challenges are still not adequately developed, resulting in insufficient improvements in the quality of management decisions and the efficiency of factory operations.

[1528] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1529] This invention includes a server that collects executives' careers, past decisions, records of statements, and related literature and generates them as a centralized dataset; a server that trains a machine learning model based on the pre-processed data to generate a digital twin of the executives; a server that collects operational data of machinery and equipment used in the factory and generates an optimized operational plan; a server that analyzes data and scenarios related to new management decisions and runs simulations; and a server that visually displays the results of the simulations to the user. This enables the real-time utilization of management insights to support management decisions and further improves and optimizes the efficiency of factory operations.

[1530] An "executive" is a person who is in a position to make important decisions within a company or organization and who possesses extensive knowledge and experience in management.

[1531] "Background" refers to information that shows what kind of jobs and roles a person has experienced in the past.

[1532] A "record of statements" is data that compiles and saves statements and comments made by a specific person in the past.

[1533] "Related literature" refers to documentary materials such as books, articles, and reports related to the person or event in question.

[1534] A "centralized dataset" is a collection of data that integrates data gathered from multiple different sources and organizes it into a manageable format.

[1535] A "machine learning model" is a data analysis algorithm created based on collected data to perform predictions and classifications.

[1536] A "digital twin" is a virtual replica of a real-world object, such as a data-based imitation of an executive's knowledge and experience.

[1537] "Machinery and equipment used in factories" refers to the machinery and equipment used in manufacturing and production processes.

[1538] "Operational data" refers to specific numerical data related to the operational status of machinery and equipment, such as operating status, production volume, and the number of errors that occurred.

[1539] An "optimized operational plan" is a plan formulated with the aim of ensuring the efficient and effective operation of machinery and equipment.

[1540] "Data related to new management decisions" refers to data that shows information and scenarios about new business strategies, investment plans, etc.

[1541] A "simulation" is a test or experiment that simulates real-world situations, and is a method for making predictions or evaluations under specific conditions.

[1542] "Means of visual display" refers to methods of displaying data and simulation results in a format that is easy for users to understand, and includes graphs, charts, dashboards, and so on.

[1543] This invention is a system for digitizing executive knowledge and supporting management decisions. The system consists of three components: a terminal, a server, and a user. A specific embodiment of this system is described below.

[1544] System Configuration

[1545] This system primarily uses the following hardware and software:

[1546] Hardware: Smartphones and tablets (devices), servers

[1547] Software: Python, Pandas, Scikit-learn, Matplotlib

[1548] Program processing

[1549] Data collection and transmission

[1550] Users input executives' career histories, past decisions, statements, and relevant literature into a terminal using a smartphone or tablet. Users also input operational data for machinery used in the factory (e.g., operating hours, number of errors, production quantity). The terminal normalizes the input data and sends it to the server as a centralized dataset.

[1551] Data preprocessing and model generation

[1552] The server receives data sent from terminals and performs preprocessing such as text normalization, tokenization, and standardization. The server then trains machine learning models based on the preprocessed data to generate digital twins of executives. Additionally, the server generates optimized operational plans using operational data from machinery used in factories and saves each model.

[1553] Simulation execution and result display

[1554] Users input data and scenarios related to new business decisions into a terminal. Examples include scenarios for new product launches or business acquisitions, annual production targets, and maintenance schedules. The terminal sends this data to a server. The server runs simulations using a digital twin model and operational data model to generate forecast results. These forecast results include risk assessments, expected profits, and production plans. The server sends the simulation results back to the terminal, which displays them visually. Display formats include graphs, charts, and dashboards.

[1555] Specific example: Optimizing the annual operational plan for factory robots.

[1556] The user inputs annual operational data (e.g., monthly production targets, maintenance schedules for various machines, etc.). The following prompts are used as an example:

[1557] Example of an input prompt:

[1558] Please enter the following operational data:

[1559] Months: January, February, March...December

[1560] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[1561] Maintenance scheduled for: January 15th, February 20th...

[1562] Starting the simulation.

[1563] This allows users to input necessary data through their terminals, and the server to run simulations using operational data models, generating and presenting optimal operational plans.

[1564] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1565] Step 1:

[1566] Users input executives' career histories, past decisions, statements, relevant literature, and operational data of machinery used in the factory into a terminal. The entered data includes executive attribute information (e.g., name, past duties), details of statements, literature citations, working hours, number of errors, and production quantities. The terminal centrally collects this data, normalizes it as needed, and corrects inconsistencies.

[1567] Step 2:

[1568] The device collects and normalizes the data, constructs it as a dataset, and sends it to the server. During this process, the data is converted to a standardized format such as CSV or JSON. The receiving server receives this data and stores it in its database.

[1569] Step 3:

[1570] The server preprocesses the received data. This includes natural language processing (NLP) techniques such as text normalization, tokenization, and stop word removal. Numerical data is also preprocessed through standardization and normalization. The preprocessed data is converted into a parseable format and proceeds to the next processing step.

[1571] Step 4:

[1572] The server trains machine learning models using pre-processed data. These models include digital twin models of executives and models that generate optimal operational plans for machinery and equipment used in factories. These models are trained using random forest regression and deep learning techniques. The generated models are stored on the server.

[1573] Step 5:

[1574] The user inputs data and scenarios related to new business decisions into the terminal. Specifically, they provide scenario information such as new product launches, business acquisitions, and new production targets. The following prompts are used:

[1575] Example of an input prompt:

[1576] Please enter the following operational data:

[1577] Months: January, February, March...December

[1578] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[1579] Maintenance scheduled for: January 15th, February 20th...

[1580] Starting the simulation.

[1581] Step 6:

[1582] The terminal sends the new data and scenario entered by the user to the server. The server receives this data, performs preprocessing and standardization, and then runs a simulation using the stored digital twin model and operational planning model.

[1583] Step 7:

[1584] The server generates the results of the executed simulation. These predictions include risk assessments, expected profits, and optimized production plans. The simulation results are stored as a dataset on the server and sent to the terminal.

[1585] Step 8:

[1586] The terminal visually displays simulation results received from the server to the user. This includes graphs, charts, and dashboards. Users can use this information to make specific business decisions and operational plan decisions.

[1587] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1588] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions, and a specific embodiment thereof is shown below.

[1589] System Configuration

[1590] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[1591] Program processing

[1592] Data collection and transmission

[1593] User

[1594] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Furthermore, users input their own emotional state or have it automatically recognized by an emotion engine.

[1595] terminal

[1596] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted for subsequent processing. The device also sends emotion recognition requests to the emotion engine.

[1597] Emotional Engine

[1598] The emotion engine analyzes the user's emotions using user input data or facial and speech recognition technologies. The analysis results are sent to a server for later use in simulations.

[1599] server

[1600] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database.

[1601] Data preprocessing and model generation

[1602] server

[1603] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[1604] server

[1605] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[1606] server

[1607] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[1608] server

[1609] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[1610] Simulation execution and result display

[1611] User

[1612] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[1613] terminal

[1614] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[1615] server

[1616] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[1617] server

[1618] The server exports the simulation results as a dataset and sends it to the terminal.

[1619] terminal

[1620] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[1621] Specific example

[1622] Example 1: Launching a new product into market

[1623] User

[1624] The user inputs information about the market launch of the new smart device (e.g., key product features, target market, sales budget) into the device. Emotional states are also input simultaneously or recognized by the emotion engine.

[1625] terminal

[1626] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[1627] server

[1628] The server runs simulations using digital twin models of executives and data from the emotion engine. It generates information such as predicted sales, risk factors, and marketing strategies.

[1629] server

[1630] The server sends the simulation results to the terminal.

[1631] terminal

[1632] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1633] Example 2: Considering a business acquisition

[1634] User

[1635] The user inputs information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal. Emotional states are also input simultaneously or recognized by an emotion engine.

[1636] terminal

[1637] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[1638] server

[1639] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[1640] server

[1641] The server sends the simulation results to the terminal.

[1642] terminal

[1643] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1644] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[1645] The following describes the processing flow.

[1646] Step 1:

[1647] User

[1648] Users input executives' career histories, past decisions, statements, and relevant literature into the terminal. If emotional states are also to be entered, they can be done using text or a dedicated emotion input interface.

[1649] Step 2:

[1650] terminal

[1651] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[1652] Step 3:

[1653] terminal

[1654] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[1655] Step 4:

[1656] terminal

[1657] The terminal sends the input data to the server. At the same time, it sends an emotion recognition request to the emotion engine.

[1658] Step 5:

[1659] Emotional Engine

[1660] The emotion engine analyzes the user's emotions using user input data, facial recognition, and speech recognition technology. The analysis results are then sent to the server.

[1661] Step 6:

[1662] server

[1663] The server receives data sent from the terminal and the emotion engine and stores it in a database. This database is used for subsequent processing.

[1664] Step 7:

[1665] server

[1666] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[1667] Step 8:

[1668] server

[1669] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[1670] Step 9:

[1671] server

[1672] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[1673] Step 10:

[1674] server

[1675] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[1676] Step 11:

[1677] User

[1678] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[1679] Step 12:

[1680] terminal

[1681] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[1682] Step 13:

[1683] server

[1684] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[1685] Step 14:

[1686] server

[1687] The server exports the simulation results as a dataset and sends it to the terminal.

[1688] Step 15:

[1689] terminal

[1690] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[1691] Step 16:

[1692] User

[1693] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[1694] (Example 2)

[1695] Next, we will describe Example 2. 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."

[1696] In recent years, it has become increasingly important for companies to consider not only the experience and knowledge of executives but also the emotions and subjectivity of users when making management decisions. However, there is a lack of systems that can properly centralize, analyze, and visually display the results of simulations based on this information. As a result, the management decision-making process is not conducted efficiently and effectively, leading to a decline in the quality of decision-making in corporate operations.

[1697] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1698] In this invention, the server includes means for inputting or recognizing the user's emotional state, means for collecting the executive's career history, past decisions, records of statements, and related literature, and for formatting and storing them as a centralized dataset, means for training a machine learning model based on the pre-processed data and generating a digital twin of the executive, means for analyzing data and scenarios related to new management decisions while considering the user's emotional state, and for running simulations, and means for exporting the simulation results as a dataset and further displaying it visually. This makes it possible to comprehensively manage and analyze diverse information and support management decisions that reflect emotions.

[1699] "User emotional state" refers to the user's emotions and psychological state, and includes information input through the human-machine interface and emotional information recognized using an emotion engine.

[1700] A "centralized dataset" is a collection of data that integrates and manages executives' career histories, past decisions, statements, and related literature, and then formats and stores them in a format suitable for analysis and processing.

[1701] "Preprocessed data" refers to data that has been transformed into a format suitable for training machine learning models by applying processing such as normalization, tokenization, stop word removal, and stemming to the input data using techniques such as natural language processing.

[1702] A "machine learning model" is an algorithm and system that learns patterns from data and uses that knowledge to make predictions, classifications, and identifications.

[1703] A "digital twin" is a digital twin model created based on machine learning models to reproduce physical executives, their actions, and decision-making processes in a virtual environment.

[1704] A "scenario" refers to a specific combination of circumstances or conditions that should be considered for business decisions, and specifically describes case studies such as the launch of a new product or the acquisition of a business.

[1705] A "simulation" is a computational process that involves trial and error in a virtual environment before actually performing an action, and uses the results to derive risk assessments and optimal strategies.

[1706] "Visual display methods" refer to ways of providing data and analysis results to users in an intuitive and easy-to-understand manner using formats such as graphs, charts, and dashboards.

[1707] A "generative AI model" is an artificial intelligence model that learns patterns and characteristics of data and makes predictions and suggestions based on new data.

[1708] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions. The system is configured as follows:

[1709] System Configuration

[1710] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[1711] Program processing

[1712] Data collection and transmission

[1713] User

[1714] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Users also input their own emotional state or undergo automatic recognition by an emotion engine.

[1715] terminal

[1716] The terminal centrally manages user input data and builds it into a dataset. The data is formatted into an appropriate format. Furthermore, the terminal sends emotion recognition requests to the emotion engine. A typical PC or mobile device can be used as the terminal.

[1717] Emotional Engine

[1718] The emotion engine analyzes user emotions using user input data, facial recognition, and speech recognition technologies. Specifically, it performs facial recognition using Python and the OpenCV library, and speech analysis using Praat. The analysis results are sent to the server in real time.

[1719] server

[1720] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL).

[1721] Data preprocessing and model generation

[1722] server

[1723] The server preprocesses the received data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming. The preprocessed data is then used to train a machine learning model using TensorFlow or PyTorch. Furthermore, the user's emotional state is also utilized as a parameter in the model.

[1724] server

[1725] The server optimizes the model parameters using a dataset to generate a digital twin for each executive. During this process, the executives' behavioral patterns and decision-making processes are learned. The optimized digital twin model is then stored in a database.

[1726] Simulation execution and result display

[1727] User

[1728] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[1729] terminal

[1730] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[1731] server

[1732] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[1733] server

[1734] The server exports the simulation results as a dataset and sends those results to the terminal.

[1735] terminal

[1736] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (graphs, charts, dashboards, etc.). Data visualization libraries such as D3.js and Chart.js are used for display. In addition, feedback information from the emotion engine is also displayed.

[1737] Specific example

[1738] Example 1: Launching a new product into market

[1739] User

[1740] Users input information about the market launch of a new smart device (key product features, target market, sales budget) into the device. Their emotional state is also input simultaneously or recognized by an emotion engine.

[1741] terminal

[1742] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[1743] server

[1744] The server runs simulations using executive digital twin models and emotion engine data, generating information on predicted sales, risk factors, and marketing strategies.

[1745] server

[1746] The server sends the simulation results to the terminal.

[1747] terminal

[1748] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1749] Example 2: Considering a business acquisition

[1750] User

[1751] The user inputs information about the target company (financial status, synergy effects, acquisition price) into the terminal. Their emotional state is also entered simultaneously or recognized by an emotion engine.

[1752] terminal

[1753] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[1754] server

[1755] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results for acquisition risk assessment, expected returns, and optimal strategies.

[1756] server

[1757] The server sends the simulation results to the terminal.

[1758] terminal

[1759] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[1760] Example of a prompt

[1761] 1. "A user is considering launching a new product. Use an emotion engine to analyze the user's emotional state and display simulation results to support their decision-making regarding market launch."

[1762] 2. "A user is considering acquiring a certain company. Based on the target company's data and the results of the sentiment engine, perform a simulation of acquisition risk and expected return, and propose the optimal strategy."

[1763] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[1764] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1765] The processing flow of a system program is divided into processing steps.

[1766] Step 1:

[1767] User

[1768] Users input data such as executives' career histories, past decisions, statements, and relevant literature into the terminal. This input includes text data and scanned documents. In addition, users can manually input their own emotional state or have it automatically recognized by an emotion engine.

[1769] Input: Executive data, literature data, emotional state

[1770] Output: Centralized dataset

[1771] Step 2:

[1772] terminal

[1773] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted into an appropriate format. The device temporarily stores the formatted data and sends an emotion recognition request to the emotion engine.

[1774] Input: User data, emotional state

[1775] Data processing: Data format conversion, data formatting.

[1776] Output: Formatted dataset, sentiment recognition requests

[1777] Step 3:

[1778] Emotional Engine

[1779] The emotion engine analyzes user input data and video and audio data obtained from the device's camera and microphone. It uses facial recognition technology and voice analysis (voice feature extraction and emotion classification) to quantify the user's emotional state. The analysis results are transmitted to the server in real time.

[1780] Input: Video data, audio data

[1781] Data processing: facial recognition, voice analysis

[1782] Output: Sentiment analysis results

[1783] Step 4:

[1784] server

[1785] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL).

[1786] Input: Formatted dataset, sentiment analysis results

[1787] Data processing: Data reception, data storage

[1788] Output: Saved database

[1789] Step 5:

[1790] server

[1791] The server preprocesses the stored data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming.

[1792] Input: Saved data

[1793] Data processing: Text preprocessing (normalization, tokenization, stop word removal, stemming)

[1794] Output: Preprocessed data

[1795] Step 6:

[1796] server

[1797] The server trains machine learning models using TensorFlow or PyTorch based on preprocessed data. Furthermore, it utilizes the user's emotional state as a parameter in the model.

[1798] Input: Preprocessed data, emotional state

[1799] Data processing: Machine learning model training

[1800] Output: Trained machine learning model

[1801] Step 7:

[1802] server

[1803] The server optimizes the parameters of the digital twin model using executive data. In this process, the executives' behavioral patterns and decision-making processes are learned.

[1804] Input: Executive data, machine learning model

[1805] Data Calculation: Model Parameter Optimization

[1806] Output: Optimized digital twin model

[1807] Step 8:

[1808] server

[1809] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[1810] Input: Optimized digital twin model

[1811] Data processing: Data storage

[1812] Output: Saved digital twin model

[1813] Step 9:

[1814] User

[1815] Users input data and scenarios related to new business decisions into the terminal. Specific examples include information on new product launches and business acquisitions. Emotional states are either entered sequentially or automatically updated by the emotion engine.

[1816] Input: New scenario data, emotional state

[1817] Output: Input scenario data

[1818] Step 10:

[1819] terminal

[1820] The device sends new scenario data and sentiment analysis results to the server.

[1821] Input: Input scenario data, sentiment analysis results

[1822] Data processing: Data transmission

[1823] Output: Sent data

[1824] Step 11:

[1825] server

[1826] The server uses information from the digital twin model and the emotion engine to perform simulations and generate prediction results. Specific examples include risk assessment, expected profit, and optimal strategy.

[1827] Input: Digital twin model, scenario data, sentiment analysis results

[1828] Data calculation: Simulation execution

[1829] Output: Simulation results

[1830] Step 12:

[1831] server

[1832] The server exports the simulation results as a dataset and sends it to the terminal.

[1833] Input: Simulation results

[1834] Data processing: Data export, data transmission

[1835] Output: Sent simulation results

[1836] Step 13:

[1837] terminal

[1838] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (graphs, charts, dashboards, etc.). It uses data visualization libraries such as D3.js and Chart.js. It also displays feedback information from the emotion engine.

[1839] Input: Simulation results, emotional feedback

[1840] Data processing: Data visualization, data display

[1841] Output: Visualized simulation results

[1842] (Application Example 2)

[1843] Next, we will explain application example 2. In the following explanation, 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."

[1844] Conventional emotion recognition technologies and business decision support systems have struggled to reflect users' emotional states in business decisions, resulting in low accuracy in providing feedback for specific business scenarios. Furthermore, the lack of real-time customer service improvements based on customer emotions in physical stores has led to inconsistencies in the quality of the customer experience.

[1845] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[1846] This invention includes a server that collects executives' careers, past decisions, statements, and related literature and generates them as a centralized dataset; a server that trains a machine learning model based on the pre-processed data to generate a digital twin of the executives; a server that analyzes data and scenarios related to new management decisions and runs simulations; a server that recognizes the emotions of customers' facial expressions and voices in real time and provides feedback on customer service methods; and a server that visually displays the results of the simulations to the user. This makes it possible to improve the accuracy of management decisions by utilizing the emotional state of the user, and further enables real-time improvement of customer service based on customer emotions in physical stores.

[1847] The term "executive" refers to a person who is in a position to make important decisions regarding the management of a company or organization.

[1848] "Career history" refers to a series of records including an executive's work history, educational history, past duties and responsibilities, and achievements.

[1849] "Past decisions" refers to records of important judgments and decisions made by executives regarding management and operations to date.

[1850] "Records of statements" refer to information that records the opinions and statements made by executives in official settings or meetings.

[1851] "Relevant literature" refers to reference materials such as books, papers, reports, and articles related to executives and management decisions.

[1852] A "centralized dataset" refers to a collection of data compiled from large amounts of information gathered from different sources into a unified format or style.

[1853] A "machine learning model" refers to an algorithm that learns patterns and features based on large amounts of data, and then uses that learning to make predictions and classifications on new data.

[1854] A "digital twin" refers to a digital model that recreates a real-world physical object or person in a virtual space.

[1855] A "scenario" refers to the setting of a simulation that summarizes specific situations and assumptions related to new management decisions or strategies.

[1856] "Simulation" refers to the process of virtually predicting the outcome of business decisions based on a set scenario.

[1857] "Visual display" refers to presenting data and analysis results to users in visual formats such as graphs, charts, and dashboards.

[1858] "Customer" refers to consumers or clients who use the products or services of a company or store.

[1859] "Emotional cues in facial expressions and voice" refers to the inner emotional state that can be gleaned from a customer's facial expressions, tone of voice, and intonation.

[1860] "Customer service methods" refer to the ways in which one responds to and acts to effectively provide services or products to customers.

[1861] "Feedback" refers to the provision of evaluations and advice regarding actions and decisions by systems or individuals.

[1862] "Real-time" refers to the processing and reaction of information occurring at roughly the same time as events in the real world.

[1863] System Overview

[1864] This invention uses a centralized dataset including executives' career histories and past decisions to generate digital twins of executives, simulate new management decisions, and visually display the results to the user. It also includes a system that recognizes the emotions of customers' facial expressions and voices in real time in physical stores and provides feedback on how to serve them.

[1865] Hardware and software configuration

[1866] Hardware:

[1867] Terminal: A digital terminal with advanced display and input devices.

[1868] Camera: High-resolution camera built into the smart glasses.

[1869] Microphone: A high-precision microphone built into the smart glasses.

[1870] software:

[1871] Emotion analysis engine: DeepFace and Google Speech-to-Text for analyzing user and customer facial expressions and voices.

[1872] Natural Language Processing (NLP): The Hugging Face transformers library is used for processing text data.

[1873] Machine learning models: PyTorch and TensorFlow are used for training based on datasets.

[1874] System operation and data processing

[1875] 1. Data Collection: Users input executives' resumes, past decisions, and relevant literature into their devices, and this information is sent to the server as a centralized dataset.

[1876] 2. Digital Twin Generation: Based on the pre-processed data, the server generates digital twins of the executives and stores them in the database. This involves the application of NLP techniques such as text normalization, tokenization, stop word removal, and stemming.

[1877] 3. Simulation: Data and scenarios related to new business decisions are entered via a terminal, and the simulation is executed on the server. This generates and visually displays the prediction results.

[1878] 4. Emotion Recognition and Feedback: In physical stores, customer emotions are analyzed in real time from facial expressions and voice, and feedback on customer service methods is displayed on the smart glasses' screen.

[1879] Specific examples

[1880] New product market launch: Users input information about the new product into a terminal, and based on simulations, they can understand predicted sales and risk factors.

[1881] Business acquisition consideration: Users submit information about potential acquisition targets to a server to obtain risk assessments and forecasts of expected returns.

[1882] Improving customer service in physical stores: Employees wearing smart glasses can recognize customer emotions in real time while interacting with customers and receive feedback on how to respond appropriately.

[1883] Example of a prompt

[1884] If a customer says, "This is a bit expensive," and looks dissatisfied:

[1885] "The customer seems dissatisfied. Please suggest other options."

[1886] When simulating the market launch of a new product:

[1887] "Input: Information regarding the market launch of new smart devices. Output: Projected sales, risk factors, and marketing strategy."

[1888] summary

[1889] This invention makes it possible to improve the accuracy of business decisions by utilizing the emotional state of users, and enables real-time improvements to customer service based on customer emotions in physical stores.

[1890] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1891] Step 1:

[1892] Users input executives' career histories, past decisions, statements, and related literature into a terminal to generate a centralized dataset. Input data includes text, scanned documents, etc. The terminal formats this data into a unified format and sends it to the server.

[1893] Step 2:

[1894] The server preprocesses the received data. Specifically, natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming are applied. The preprocessed data is then converted into a format suitable for training machine learning models.

[1895] Step 3:

[1896] The server trains a machine learning model based on pre-processed data to generate a digital twin of the executives. Deep learning and reinforcement learning algorithms are used to train the model. The generated digital twin model learns the executives' behavioral patterns and decision-making processes.

[1897] Step 4:

[1898] The server stores the trained digital twin model in a database, making it available for simulation in new scenarios. The stored model can be quickly accessed for subsequent processing.

[1899] Step 5:

[1900] The user inputs data and scenarios related to new business decisions into the device. This may include scenario information for new product launches or business acquisitions. The user's emotional state is also recognized either through input or automatically by the emotion engine.

[1901] Step 6:

[1902] The terminal sends the newly entered data and scenario, along with the results of the emotion engine's analysis, to the server. The transmitted data forms the basis for the new simulation.

[1903] Step 7:

[1904] The server combines the digital twin model with new data to run simulations. This includes risk assessment, expected returns, and optimal strategies. The results are generated in a format that is visually displayed to the user.

[1905] Step 8:

[1906] The server sends the simulation results to the terminal. The terminal generates graphs, charts, and dashboards to visually display the received results in an easy-to-understand manner.

[1907] Step 9:

[1908] In physical stores, the smart glasses' camera and microphone are used to recognize customers' facial expressions and voices in real time. The emotion engine analyzes the video and audio data to identify the customer's emotions.

[1909] Step 10:

[1910] Based on the emotion analysis results, the server generates appropriate customer service feedback and displays it on the smart glasses' display. If the customer is happy, it will provide specific instructions such as "Please continue to assist them," while if they are dissatisfied, it will provide instructions such as "Please try offering a different suggestion."

[1911] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1912] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1913] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1914] [Fourth Embodiment]

[1915] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1916] As shown in Figure 7, the 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.

[1917] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1918] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1919] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1920] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1921] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1922] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive 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 robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1923] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1924] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[1925] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1926] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1927] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1928] This invention is a system for digitizing executives' expertise and supporting management decisions, and a specific embodiment thereof is shown below.

[1929] System Configuration

[1930] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[1931] Program processing

[1932] Data collection and transmission

[1933] User

[1934] Users input data such as executives' resumes, past decisions, statements, and relevant literature into the terminal.

[1935] Users collect information from external data sources and upload it to their devices.

[1936] terminal

[1937] The terminal centrally manages the input data and builds a dataset for transmission to the server.

[1938] The terminal standardizes the data format and corrects inconsistencies.

[1939] server

[1940] The server receives data sent from the terminal and stores it in the database.

[1941] Data preprocessing and model generation

[1942] server

[1943] The server preprocesses the received data using natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[1944] The server trains a machine learning model based on pre-processed data to generate a digital twin of the executives.

[1945] The server saves the generated digital twin model to a database.

[1946] Simulation execution and result display

[1947] User

[1948] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[1949] terminal

[1950] The terminal sends the entered data to the server.

[1951] server

[1952] The server runs simulations using a digital twin model and generates prediction results. These simulation results include risk assessments, expected profits, and optimal strategies.

[1953] The server exports the simulation results as a dataset and sends it to the terminal.

[1954] terminal

[1955] The terminal visually displays the simulation results received from the server. This can be done using formats such as graphs, charts, and dashboards.

[1956] Specific example

[1957] Example 1: Launching a new product into market

[1958] User

[1959] The user enters information about the market launch of a new smart device model into the terminal (e.g., key product features, target market, sales budget).

[1960] terminal

[1961] The terminal normalizes the input data and sends it to the server as a dataset.

[1962] server

[1963] The server runs simulations using digital twin models of executives, generating information such as projected sales, risk factors, and marketing strategies.

[1964] The server sends the simulation results to the terminal.

[1965] terminal

[1966] The terminal visually displays the simulation results received from the server to the user.

[1967] Example 2: Considering a business acquisition

[1968] User

[1969] The user enters information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal.

[1970] terminal

[1971] The terminal sends the input data to the server as a dataset.

[1972] server

[1973] The server runs simulations using digital twin models of executives. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[1974] The server sends the simulation results to the terminal.

[1975] terminal

[1976] The terminal visually displays the simulation results received from the server to the user.

[1977] This system digitizes the knowledge and experience of executives and utilizes it in real time for management decisions, thereby supporting companies in effectively formulating long-term strategies.

[1978] The following describes the processing flow.

[1979] Step 1:

[1980] User

[1981] Users input executives' career histories, past decisions, statements, and relevant documents into the terminal. This includes text data and scanned documents.

[1982] Step 2:

[1983] terminal

[1984] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[1985] Step 3:

[1986] terminal

[1987] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[1988] Step 4:

[1989] terminal

[1990] The terminal sends the centralized dataset to the server. During this process, it detects data transfer errors and attempts to resend the data if necessary.

[1991] Step 5:

[1992] server

[1993] The server receives data sent from the terminal and stores it in a database. This database is used for subsequent processing.

[1994] Step 6:

[1995] server

[1996] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[1997] Step 7:

[1998] server

[1999] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning.

[2000] Step 8:

[2001] server

[2002] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[2003] Step 9:

[2004] server

[2005] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[2006] Step 10:

[2007] User

[2008] Users input data and scenarios related to new business decisions into the terminal. For example, they might input scenarios for launching a new product or acquiring a business.

[2009] Step 11:

[2010] terminal

[2011] The terminal sends the new data and scenario entered by the user to the server.

[2012] Step 12:

[2013] server

[2014] The server analyzes new data using a digital twin model and runs simulations. The simulation results include risk assessments, expected profits, and optimal strategies.

[2015] Step 13:

[2016] server

[2017] The server exports the simulation results as a dataset and sends it to the terminal.

[2018] Step 14:

[2019] terminal

[2020] The terminal displays the simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards, etc.).

[2021] Step 15:

[2022] User

[2023] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[2024] (Example 1)

[2025] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[2026] Traditional management decision-making systems struggle to digitize the knowledge and experience of executives and utilize it in real time, making it difficult to improve the quality of management decisions. Furthermore, there is a need for a system that can centrally manage and effectively analyze data on executives' past decisions and statements.

[2027] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[2028] This invention includes a server that collects the career history, past decisions, statements, and related literature of executives and generates them as a centralized dataset; a server that normalizes, tokenizes, removes stop words, and stems the data using natural language processing techniques based on the preprocessed data; and a server that trains a machine learning model based on the preprocessed data to generate a digital twin of the executives. This makes it possible to digitize the knowledge and experience of executives and utilize it in management decisions in real time.

[2029] "Executives" refer to individuals who make important decisions within a company or organization, and primarily include management and executives.

[2030] "Career history" refers to the historical information of an employee, including their past job duties, achievements, and educational background.

[2031] "Past decisions" refers to specific decisions, judgments, or statements of intent made by an executive in the past.

[2032] "Record of statements" refers to data in text format that records statements and comments made by officials in the past.

[2033] "Literature" refers to materials and information sources such as books, papers, and business reports related to the person in charge.

[2034] A "dataset" refers to a collection of data that has been centrally managed and organized for use in analysis and model training.

[2035] "Natural language processing technology" refers to techniques for converting text data into a format that is easily understood by machines, and includes normalization, tokenization, stop word removal, stemming, and other methods.

[2036] "Normalization" refers to the process of handling character size, spaces, and special characters in order to unify the format of data.

[2037] "Tokenization" is the process of dividing text into words and phrases, which means breaking down a sentence into units that are easier to analyze.

[2038] "Stop word removal" refers to the process of excluding certain common words (such as "a" or "the") from the analysis.

[2039] "Stemming" is the process of extracting only the stem portion of a word, and it refers to unifying the inflected forms of the word.

[2040] A "machine learning model" refers to an algorithm that learns patterns and regularities based on data and uses that knowledge to make predictions and classifications on new data.

[2041] A "digital twin" is a digital representation that mimics the knowledge and judgment of real-world officials, and is used for simulations and predictions.

[2042] A "scenario" refers to a specific situation or plan related to a new management decision.

[2043] "Simulation" refers to using digital twin models to make predictions and analyses about new situations and scenarios.

[2044] "Visualization" refers to techniques that visually display simulation results using graphs, charts, and other methods to make them easier to understand.

[2045] This invention is a system for digitizing the knowledge of executives and supporting management decisions. Its specific embodiments are as follows:

[2046] System Configuration

[2047] This system consists of three main components: terminals, servers, and users. Terminals handle data input and output, while servers perform data processing and simulations. Users operate the system and utilize it for business decision-making.

[2048] Data collection and transmission

[2049] User

[2050] Users input data such as the career history of executives, past decisions, records of statements, and relevant literature into spreadsheets or dedicated forms. They also collect information from external data sources using APIs and scraping tools and upload it to their devices.

[2051] terminal

[2052] The terminal converts the input data into a CSV file and builds a dataset for centralized management. It standardizes the data format using Excel or Google Sheets, detects and corrects inconsistent data, and sends the corrected data to the server.

[2053] server

[2054] The server receives data sent from the terminal and saves it to a dedicated database (e.g., PostgreSQL). During this process, it checks the data's integrity to ensure there is no invalid data.

[2055] Data preprocessing

[2056] server

[2057] The server uses natural language processing techniques (e.g., Python's NLP library spaCy) to normalize, tokenize, remove stop words from, and stem the stored data. For example, all characters are converted to lowercase, unnecessary spaces and special characters are removed, and then tokenization is performed. This divides the data into units of words and phrases, common words (stop words) are removed, and stemming is performed to extract only the word stems.

[2058] Generation of a digital twin model

[2059] server

[2060] The server trains a digital twin model using machine learning algorithms (e.g., TensorFlow, scikit-learn) based on preprocessed data. For example, it uses random forests or neural networks to generate a digital twin, which is a digital representation that mimics the knowledge of the person in charge. The trained model is stored in a dedicated database for performance evaluation and optimization.

[2061] Running the simulation and displaying the results.

[2062] User

[2063] The user inputs data and scenarios related to new business decisions into the terminal. For example, they might input a specific scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[2064] terminal

[2065] The terminal converts the entered scenario data into JSON format and sends it to the server.

[2066] server

[2067] The server uses a digital twin model to perform scenario-based simulations and generates predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising" and send them to the terminal.

[2068] terminal

[2069] The terminal visually displays the received simulation results. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[2070] Specific example

[2071] Example 1: Launching a new product into market

[2072] User

[2073] The user inputs information about the market launch of a new smart device model into the terminal. For example, they might input, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[2074] terminal

[2075] The terminal converts the input data into JSON format and sends it to the server.

[2076] server

[2077] The server runs a simulation using a digital twin model, generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," and sends them to the terminal.

[2078] terminal

[2079] The terminal displays the received results in graphs. Sales forecasts are shown as bar graphs, risk assessments as heatmaps, and recommended strategies as lists.

[2080] Example 2: Considering a business acquisition

[2081] User

[2082] The user inputs the following data into the terminal: "Company Z is being considered for acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[2083] terminal

[2084] The terminal converts this data into a CSV file and sends it to the server.

[2085] server

[2086] The server runs a simulation using a digital twin model, generates results such as "Risk assessment: High, Expected return: 2 billion yen, Recommended strategy: Enhanced due diligence," and sends them to the terminal.

[2087] terminal

[2088] The terminal displays the received results in charts. Risk assessment is shown as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[2089] This will enable the digitization of executives' knowledge and experience, allowing them to be used in real-time management decisions. The system will support companies in effectively formulating long-term strategies.

[2090] The flow of the specific processing in Example 1 will be explained using Figure 11.

[2091] Step 1: Data collection and input

[2092] User

[2093] Users input data such as the career history, past decisions, statements, and relevant literature of executives into spreadsheets or dedicated forms. For example, a user might input "Executive A's career history, past decisions, and relevant literature" into a spreadsheet. They can also collect information from external data sources using APIs or scraping tools and upload it to their device. The input data is saved in CSV or Excel format.

[2094] terminal

[2095] The terminal receives the input data, converts it into a CSV file, and builds a centrally managed dataset. For example, it standardizes the data format using Excel or Google Sheets, detects inconsistent data, and corrects it. The corrected data is prepared as a temporary storage file.

[2096] Input: Records of the person's career history, past decisions, statements, and relevant literature (CSV format file)

[2097] Output: Centrally managed dataset (CSV format file)

[2098] Step 2: Send

[2099] terminal

[2100] The terminal sends a centrally managed dataset to the server. During this process, the integrity of the files is checked, and retransmission is performed if necessary. For example, a data transmission API is used to send the dataset to the server.

[2101] Input: Centrally managed dataset (CSV format file)

[2102] Output: Request to send data to the server

[2103] server

[2104] The server receives data sent from the terminal and stores it in a database. For example, it might store it in a PostgreSQL database and perform data integrity checks.

[2105] Input: Data transmission request

[2106] Output: Data stored in the database

[2107] Step 3: Data Preprocessing

[2108] server

[2109] The server preprocesses the stored data using natural language processing techniques. First, it performs text normalization, converting all characters to lowercase and removing unnecessary spaces and special characters. Next, it performs tokenization, dividing the data into units of words and phrases. Furthermore, it removes stop words and extracts the stem portion of words using stemming.

[2110] Input: Data stored in the database

[2111] Output: Preprocessed text data

[2112] Specific operation: Use the Python spaCy library to perform text normalization, tokenization, stop word removal, and stemming.

[2113] Step 4: Generating a Digital Twin Model

[2114] server

[2115] The server trains a digital twin model using machine learning algorithms based on pre-processed data. Specifically, it uses random forest and neural network algorithms to generate a digital twin that mimics the knowledge of the executives. The trained model undergoes performance evaluation and optimization before being stored in a database.

[2116] Input: Preprocessed text data

[2117] Output: Trained digital twin model

[2118] Specific actions: Train the model using TensorFlow or scikit-learn.

[2119] Step 5: Run the simulation

[2120] User

[2121] The user inputs a scenario related to a new business decision into the terminal. For example, they might input a scenario such as, "Market launch of new product Y, target market: men in their 20s, sales budget: 30 million yen."

[2122] terminal

[2123] The terminal converts the entered scenario data into JSON format and sends it to the server.

[2124] Input: Scenario (text format)

[2125] Output: Scenario data in JSON format

[2126] server

[2127] The server uses a digital twin model to perform scenario-based simulations and generate predictive results such as risk assessment, expected profit, and optimal strategy. For example, it might generate results like "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising."

[2128] Input: Scenario data in JSON format

[2129] Output: Simulation results (JSON format)

[2130] Step 6: Displaying the simulation results

[2131] terminal

[2132] The terminal visually displays the simulation results received from the server. This uses data visualization tools such as D3.js and Tableau. For example, sales forecasts are displayed as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[2133] Input: Simulation results (JSON format)

[2134] Output: Visualized simulation results (graphs, heatmaps, lists)

[2135] Specific example

[2136] Example 1: Launching a new product into market

[2137] User

[2138] The user enters the following information into the terminal: "New product Y to be launched, target market: men in their 20s, sales budget: 30 million yen."

[2139] terminal

[2140] The device normalizes this information, converts it to JSON format, and sends it to the server.

[2141] server

[2142] The server runs a simulation using a digital twin model and generates results such as "Sales forecast: 50 million yen, Risk: Medium, Recommended marketing strategy: Strengthen social media advertising," which it then sends to the terminal.

[2143] terminal

[2144] The device visually displays the results, showing sales forecasts as bar graphs, risk assessments as heatmaps, and recommended strategies as dropdown lists.

[2145] Example 2: Considering a business acquisition

[2146] User

[2147] The user enters the following information into the terminal: "Company Z is considering acquisition; financial status: annual sales of 40 billion yen, net profit of 3 billion yen; synergy effect: integration of supply chains; acquisition price: 15 billion yen."

[2148] terminal

[2149] The terminal normalizes this information, converts it to CSV format, and sends it to the server.

[2150] server

[2151] The server runs a simulation using a digital twin model and generates results such as "Risk Assessment: High, Expected Return: 2 billion yen, Recommended Strategy: Enhanced Due Diligence," which it then sends to the terminal.

[2152] terminal

[2153] The device visually displays the results, showing risk assessments as a heatmap, expected returns as a bar graph, and recommended strategies as a list.

[2154] Following these specific processing steps, the system can digitize the knowledge and experience of executives and utilize it in real time for management decisions.

[2155] (Application Example 1)

[2156] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[2157] In modern corporate management, there is a need to leverage the insights of executives and management in real time to effectively support management decisions. Furthermore, it is crucial to efficiently collect operational data from machinery and equipment used in factories and formulate optimal operational plans. However, systems to address these challenges are still not adequately developed, resulting in insufficient improvements in the quality of management decisions and the efficiency of factory operations.

[2158] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[2159] This invention includes a server that collects executives' careers, past decisions, records of statements, and related literature and generates them as a centralized dataset; a server that trains a machine learning model based on the pre-processed data to generate a digital twin of the executives; a server that collects operational data of machinery and equipment used in the factory and generates an optimized operational plan; a server that analyzes data and scenarios related to new management decisions and runs simulations; and a server that visually displays the results of the simulations to the user. This enables the real-time utilization of management insights to support management decisions and further improves and optimizes the efficiency of factory operations.

[2160] An "executive" is a person who is in a position to make important decisions within a company or organization and who possesses extensive knowledge and experience in management.

[2161] "Background" refers to information that shows what kind of jobs and roles a person has experienced in the past.

[2162] A "record of statements" is data that compiles and saves statements and comments made by a specific person in the past.

[2163] "Related literature" refers to documentary materials such as books, articles, and reports related to the person or event in question.

[2164] A "centralized dataset" is a collection of data that integrates data gathered from multiple different sources and organizes it into a manageable format.

[2165] A "machine learning model" is a data analysis algorithm created based on collected data to perform predictions and classifications.

[2166] A "digital twin" is a virtual replica of a real-world object, such as a data-based imitation of an executive's knowledge and experience.

[2167] "Machinery and equipment used in factories" refers to the machinery and equipment used in manufacturing and production processes.

[2168] "Operational data" refers to specific numerical data related to the operational status of machinery and equipment, such as operating status, production volume, and the number of errors that occurred.

[2169] An "optimized operational plan" is a plan formulated with the aim of ensuring the efficient and effective operation of machinery and equipment.

[2170] "Data related to new management decisions" refers to data that shows information and scenarios about new business strategies, investment plans, etc.

[2171] A "simulation" is a test or experiment that simulates real-world situations, and is a method for making predictions or evaluations under specific conditions.

[2172] "Means of visual display" refers to methods of displaying data and simulation results in a format that is easy for users to understand, and includes graphs, charts, dashboards, and so on.

[2173] This invention is a system for digitizing executive knowledge and supporting management decisions. The system consists of three components: a terminal, a server, and a user. A specific embodiment of this system is described below.

[2174] System Configuration

[2175] This system primarily uses the following hardware and software:

[2176] Hardware: Smartphones and tablets (devices), servers

[2177] Software: Python, Pandas, Scikit-learn, Matplotlib

[2178] Program processing

[2179] Data collection and transmission

[2180] Users input executives' career histories, past decisions, statements, and relevant literature into a terminal using a smartphone or tablet. Users also input operational data for machinery used in the factory (e.g., operating hours, number of errors, production quantity). The terminal normalizes the input data and sends it to the server as a centralized dataset.

[2181] Data preprocessing and model generation

[2182] The server receives data sent from terminals and performs preprocessing such as text normalization, tokenization, and standardization. The server then trains machine learning models based on the preprocessed data to generate digital twins of executives. Additionally, the server generates optimized operational plans using operational data from machinery used in factories and saves each model.

[2183] Simulation execution and result display

[2184] Users input data and scenarios related to new business decisions into a terminal. Examples include scenarios for new product launches or business acquisitions, annual production targets, and maintenance schedules. The terminal sends this data to a server. The server runs simulations using a digital twin model and operational data model to generate forecast results. These forecast results include risk assessments, expected profits, and production plans. The server sends the simulation results back to the terminal, which displays them visually. Display formats include graphs, charts, and dashboards.

[2185] Specific example: Optimizing the annual operational plan for factory robots.

[2186] The user inputs annual operational data (e.g., monthly production targets, maintenance schedules for various machines, etc.). The following prompts are used as an example:

[2187] Example of an input prompt:

[2188] Please enter the following operational data:

[2189] Months: January, February, March...December

[2190] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[2191] Maintenance scheduled for: January 15th, February 20th...

[2192] Starting the simulation.

[2193] This allows users to input necessary data through their terminals, and the server to run simulations using operational data models, generating and presenting optimal operational plans.

[2194] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[2195] Step 1:

[2196] Users input executives' career histories, past decisions, statements, relevant literature, and operational data of machinery used in the factory into a terminal. The entered data includes executive attribute information (e.g., name, past duties), details of statements, literature citations, working hours, number of errors, and production quantities. The terminal centrally collects this data, normalizes it as needed, and corrects inconsistencies.

[2197] Step 2:

[2198] The device collects and normalizes the data, constructs it as a dataset, and sends it to the server. During this process, the data is converted to a standardized format such as CSV or JSON. The receiving server receives this data and stores it in its database.

[2199] Step 3:

[2200] The server preprocesses the received data. This includes natural language processing (NLP) techniques such as text normalization, tokenization, and stop word removal. Numerical data is also preprocessed through standardization and normalization. The preprocessed data is converted into a parseable format and proceeds to the next processing step.

[2201] Step 4:

[2202] The server trains machine learning models using pre-processed data. These models include digital twin models of executives and models that generate optimal operational plans for machinery and equipment used in factories. These models are trained using random forest regression and deep learning techniques. The generated models are stored on the server.

[2203] Step 5:

[2204] The user inputs data and scenarios related to new business decisions into the terminal. Specifically, they provide scenario information such as new product launches, business acquisitions, and new production targets. The following prompts are used:

[2205] Example of an input prompt:

[2206] Please enter the following operational data:

[2207] Months: January, February, March...December

[2208] Production targets (in millions): 10, 12, 11, 15, 14, 13, 12, 16, 17, 14, 15, 18

[2209] Maintenance scheduled for: January 15th, February 20th...

[2210] Starting the simulation.

[2211] Step 6:

[2212] The terminal sends the new data and scenario entered by the user to the server. The server receives this data, performs preprocessing and standardization, and then runs a simulation using the stored digital twin model and operational planning model.

[2213] Step 7:

[2214] The server generates the results of the executed simulation. These predictions include risk assessments, expected profits, and optimized production plans. The simulation results are stored as a dataset on the server and sent to the terminal.

[2215] Step 8:

[2216] The terminal visually displays simulation results received from the server to the user. This includes graphs, charts, and dashboards. Users can use this information to make specific business decisions and operational plan decisions.

[2217] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[2218] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions, and a specific embodiment thereof is shown below.

[2219] System Configuration

[2220] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[2221] Program processing

[2222] Data collection and transmission

[2223] User

[2224] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Furthermore, users input their own emotional state or have it automatically recognized by an emotion engine.

[2225] terminal

[2226] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted for subsequent processing. The device also sends emotion recognition requests to the emotion engine.

[2227] Emotional Engine

[2228] The emotion engine analyzes the user's emotions using user input data or facial and speech recognition technologies. The analysis results are sent to a server for later use in simulations.

[2229] server

[2230] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database.

[2231] Data preprocessing and model generation

[2232] server

[2233] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[2234] server

[2235] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[2236] server

[2237] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[2238] server

[2239] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[2240] Simulation execution and result display

[2241] User

[2242] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[2243] terminal

[2244] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[2245] server

[2246] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[2247] server

[2248] The server exports the simulation results as a dataset and sends it to the terminal.

[2249] terminal

[2250] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[2251] Specific example

[2252] Example 1: Launching a new product into market

[2253] User

[2254] The user inputs information about the market launch of the new smart device (e.g., key product features, target market, sales budget) into the device. Emotional states are also input simultaneously or recognized by the emotion engine.

[2255] terminal

[2256] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[2257] server

[2258] The server runs simulations using digital twin models of executives and data from the emotion engine. It generates information such as predicted sales, risk factors, and marketing strategies.

[2259] server

[2260] The server sends the simulation results to the terminal.

[2261] terminal

[2262] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[2263] Example 2: Considering a business acquisition

[2264] User

[2265] The user inputs information about the target company (e.g., financial status, synergy effects, acquisition price) into the terminal. Emotional states are also input simultaneously or recognized by an emotion engine.

[2266] terminal

[2267] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[2268] server

[2269] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results that include acquisition risk assessments, expected returns, and optimal strategies.

[2270] server

[2271] The server sends the simulation results to the terminal.

[2272] terminal

[2273] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[2274] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[2275] The following describes the processing flow.

[2276] Step 1:

[2277] User

[2278] Users input executives' career histories, past decisions, statements, and relevant literature into the terminal. If emotional states are also to be entered, they can be done using text or a dedicated emotion input interface.

[2279] Step 2:

[2280] terminal

[2281] The terminal centrally manages the data entered by the user and constructs it as a dataset. This dataset is then formatted into a format suitable for subsequent processing.

[2282] Step 3:

[2283] terminal

[2284] The terminal will standardize data formats and correct inconsistencies. For example, it will standardize date formats and number formats.

[2285] Step 4:

[2286] terminal

[2287] The terminal sends the input data to the server. At the same time, it sends an emotion recognition request to the emotion engine.

[2288] Step 5:

[2289] Emotional Engine

[2290] The emotion engine analyzes the user's emotions using user input data, facial recognition, and speech recognition technology. The analysis results are then sent to the server.

[2291] Step 6:

[2292] server

[2293] The server receives data sent from the terminal and the emotion engine and stores it in a database. This database is used for subsequent processing.

[2294] Step 7:

[2295] server

[2296] The server preprocesses the received data. Specifically, it applies natural language processing (NLP) techniques such as text normalization, tokenization, stop word removal, and stemming.

[2297] Step 8:

[2298] server

[2299] The server trains machine learning models based on pre-processed data. The algorithms used include deep learning and reinforcement learning. Furthermore, the user's emotional state is also utilized as a parameter for the model.

[2300] Step 9:

[2301] server

[2302] The server optimizes the model parameters using data for each executive to generate a digital twin of each executive. In this process, the executives' behavioral patterns and decision-making processes are learned.

[2303] Step 10:

[2304] server

[2305] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[2306] Step 11:

[2307] User

[2308] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[2309] Step 12:

[2310] terminal

[2311] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[2312] Step 13:

[2313] server

[2314] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[2315] Step 14:

[2316] server

[2317] The server exports the simulation results as a dataset and sends it to the terminal.

[2318] Step 15:

[2319] terminal

[2320] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (e.g., graphs, charts, dashboards). Furthermore, feedback information from the emotion engine is also displayed.

[2321] Step 16:

[2322] User

[2323] Users make actual business decisions based on the displayed simulation results. They can also input additional scenarios and run the simulation again as needed.

[2324] (Example 2)

[2325] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[2326] In recent years, it has become increasingly important for companies to consider not only the experience and knowledge of executives but also the emotions and subjectivity of users when making management decisions. However, there is a lack of systems that can properly centralize, analyze, and visually display the results of simulations based on this information. As a result, the management decision-making process is not conducted efficiently and effectively, leading to a decline in the quality of decision-making in corporate operations.

[2327] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[2328] In this invention, the server includes means for inputting or recognizing the user's emotional state, means for collecting the executive's career history, past decisions, records of statements, and related literature, and for formatting and storing them as a centralized dataset, means for training a machine learning model based on the pre-processed data and generating a digital twin of the executive, means for analyzing data and scenarios related to new management decisions while considering the user's emotional state, and for running simulations, and means for exporting the simulation results as a dataset and further displaying it visually. This makes it possible to comprehensively manage and analyze diverse information and support management decisions that reflect emotions.

[2329] "User emotional state" refers to the user's emotions and psychological state, and includes information input through the human-machine interface and emotional information recognized using an emotion engine.

[2330] A "centralized dataset" is a collection of data that integrates and manages executives' career histories, past decisions, statements, and related literature, and then formats and stores them in a format suitable for analysis and processing.

[2331] "Preprocessed data" refers to data that has been transformed into a format suitable for training machine learning models by applying processing such as normalization, tokenization, stop word removal, and stemming to the input data using techniques such as natural language processing.

[2332] A "machine learning model" is an algorithm and system that learns patterns from data and uses that knowledge to make predictions, classifications, and identifications.

[2333] A "digital twin" is a digital twin model created based on machine learning models to reproduce physical executives, their actions, and decision-making processes in a virtual environment.

[2334] A "scenario" refers to a specific combination of circumstances or conditions that should be considered for business decisions, and specifically describes case studies such as the launch of a new product or the acquisition of a business.

[2335] A "simulation" is a computational process that involves trial and error in a virtual environment before actually performing an action, and uses the results to derive risk assessments and optimal strategies.

[2336] "Visual display methods" refer to ways of providing data and analysis results to users in an intuitive and easy-to-understand manner using formats such as graphs, charts, and dashboards.

[2337] A "generative AI model" is an artificial intelligence model that learns patterns and characteristics of data and makes predictions and suggestions based on new data.

[2338] This invention is a system that combines an emotion engine that recognizes user emotions and supports business decisions. The system is configured as follows:

[2339] System Configuration

[2340] This system consists of four main components: a terminal, a server, an emotion engine, and a user. The terminal handles data input and output, while the server processes and simulates the data. The emotion engine analyzes the user's input data to understand their emotions and provides feedback as needed. The user interacts with the system and uses it to inform business decisions.

[2341] Program processing

[2342] Data collection and transmission

[2343] User

[2344] Users input data into the terminal, including executives' career histories, past decisions, statements, and relevant literature. This includes text data and scanned documents. Users also input their own emotional state or undergo automatic recognition by an emotion engine.

[2345] terminal

[2346] The terminal centrally manages user input data and builds it into a dataset. The data is formatted into an appropriate format. Furthermore, the terminal sends emotion recognition requests to the emotion engine. A typical PC or mobile device can be used as the terminal.

[2347] Emotional Engine

[2348] The emotion engine analyzes user emotions using user input data, facial recognition, and speech recognition technologies. Specifically, it performs facial recognition using Python and the OpenCV library, and speech analysis using Praat. The analysis results are sent to the server in real time.

[2349] server

[2350] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL).

[2351] Data preprocessing and model generation

[2352] server

[2353] The server preprocesses the received data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming. The preprocessed data is then used to train a machine learning model using TensorFlow or PyTorch. Furthermore, the user's emotional state is also utilized as a parameter in the model.

[2354] server

[2355] The server optimizes the model parameters using a dataset to generate a digital twin for each executive. During this process, the executives' behavioral patterns and decision-making processes are learned. The optimized digital twin model is then stored in a database.

[2356] Simulation execution and result display

[2357] User

[2358] Users input data and scenarios related to new business decisions into the terminal. For example, they input scenarios for launching a new product or acquiring a business. Emotional states are also entered sequentially or automatically updated by the emotion engine.

[2359] terminal

[2360] The device sends new data and scenarios, as well as sentiment analysis results from the sentiment engine, to the server.

[2361] server

[2362] The server runs simulations using information from the digital twin model and the emotion engine to generate prediction results. These simulation results include risk assessment, expected profit, and optimal strategy.

[2363] server

[2364] The server exports the simulation results as a dataset and sends those results to the terminal.

[2365] terminal

[2366] The terminal displays simulation results received from the server to the user in a visually easy-to-understand format (graphs, charts, dashboards, etc.). Data visualization libraries such as D3.js and Chart.js are used for display. In addition, feedback information from the emotion engine is also displayed.

[2367] Specific example

[2368] Example 1: Launching a new product into market

[2369] User

[2370] Users input information about the market launch of a new smart device (key product features, target market, sales budget) into the device. Their emotional state is also input simultaneously or recognized by an emotion engine.

[2371] terminal

[2372] The terminal normalizes the input data and sends it to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[2373] server

[2374] The server runs simulations using executive digital twin models and emotion engine data, generating information on predicted sales, risk factors, and marketing strategies.

[2375] server

[2376] The server sends the simulation results to the terminal.

[2377] terminal

[2378] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[2379] Example 2: Considering a business acquisition

[2380] User

[2381] The user inputs information about the target company (financial status, synergy effects, acquisition price) into the terminal. Their emotional state is also entered simultaneously or recognized by an emotion engine.

[2382] terminal

[2383] The terminal sends the input data to the server as a dataset. The analysis results from the emotion engine are also sent at the same time.

[2384] server

[2385] The server runs simulations using digital twin models of executives and data from an emotion engine. It generates simulation results for acquisition risk assessment, expected returns, and optimal strategies.

[2386] server

[2387] The server sends the simulation results to the terminal.

[2388] terminal

[2389] The terminal visually displays the simulation results received from the server and the feedback from the emotion engine to the user.

[2390] Example of a prompt

[2391] 1. "A user is considering launching a new product. Use an emotion engine to analyze the user's emotional state and display simulation results to support their decision-making regarding market launch."

[2392] 2. "A user is considering acquiring a certain company. Based on the target company's data and the results of the sentiment engine, perform a simulation of acquisition risk and expected return, and propose the optimal strategy."

[2393] This system digitizes the knowledge and experience of executives and, by adding user sentiment analysis, can support more precise and human-centered management decisions.

[2394] The flow of the specific processing in Example 2 will be explained using Figure 13.

[2395] The processing flow of a system program is divided into processing steps.

[2396] Step 1:

[2397] User

[2398] Users input data such as executives' career histories, past decisions, statements, and relevant literature into the terminal. This input includes text data and scanned documents. In addition, users can manually input their own emotional state or have it automatically recognized by an emotion engine.

[2399] Input: Executive data, literature data, emotional state

[2400] Output: Centralized dataset

[2401] Step 2:

[2402] terminal

[2403] The device centrally manages the data entered by the user and builds it into a dataset. This dataset is then formatted into an appropriate format. The device temporarily stores the formatted data and sends an emotion recognition request to the emotion engine.

[2404] Input: User data, emotional state

[2405] Data processing: Data format conversion, data formatting.

[2406] Output: Formatted dataset, sentiment recognition requests

[2407] Step 3:

[2408] Emotional Engine

[2409] The emotion engine analyzes user input data and video and audio data obtained from the device's camera and microphone. It uses facial recognition technology and voice analysis (voice feature extraction and emotion classification) to quantify the user's emotional state. The analysis results are transmitted to the server in real time.

[2410] Input: Video data, audio data

[2411] Data processing: facial recognition, voice analysis

[2412] Output: Sentiment analysis results

[2413] Step 4:

[2414] server

[2415] The server receives data sent from the terminal and analysis results from the emotion engine, and stores them in a database (e.g., MySQL).

[2416] Input: Formatted dataset, sentiment analysis results

[2417] Data processing: Data reception, data storage

[2418] Output: Saved database

[2419] Step 5:

[2420] server

[2421] The server preprocesses the stored data. Specifically, it uses the Python NLTK library to perform text normalization, tokenization, stop word removal, and stemming.

[2422] Input: Saved data

[2423] Data processing: Text preprocessing (normalization, tokenization, stop word removal, stemming)

[2424] Output: Preprocessed data

[2425] Step 6:

[2426] server

[2427] The server trains machine learning models using TensorFlow or PyTorch based on preprocessed data. Furthermore, it utilizes the user's emotional state as a parameter in the model.

[2428] Input: Preprocessed data, emotional state

[2429] Data processing: Machine learning model training

[2430] Output: Trained machine learning model

[2431] Step 7:

[2432] server

[2433] The server optimizes the parameters of the digital twin model using executive data. In this process, the executives' behavioral patterns and decision-making processes are learned.

[2434] Input: Executive data, machine learning model

[2435] Data Calculation: Model Parameter Optimization

[2436] Output: Optimized digital twin model

[2437] Step 8:

[2438] server

[2439] The server saves the generated digital twin model to a database, making it available for use in subsequent simulations.

[2440] Input: Optimized digital twin model

[2441] Data processing: Data storage

[2442] Output: Saved digital twin model

[2443] Step 9:

[2444] User

[2445] Users input data and scenarios related to new business decisions into the terminal. Specific examples include information on new product launches and business acquisitions. Emotional states are either entered sequentially or automatically updated by the emotion engine.

[2446] Input: New scenario data, emotional state

[2447] Output: Input scenario data

[2448] Step 10:

[2449] terminal

[2450] The device sends new scenario data and sentiment analysis results to the server.

[2451] Input: Input scenario data, sentiment analysis results

[2452] Data processing: Data transmission

[2453] Output: Sent data

[2454] Step 11:

[2455] server

[2456] The server uses information from the digital twin model and the emotion engine to perform simulations and generate prediction results. Specific examples include risk assessment, expected profit, and optimal strategy.

[2457] Input: Digital twin model, scenario data, sentiment analysis results

[2458] Data calculation: Simulation execution

[2459] Output: Simulation results

[2460] Step 12:

[2461] server

[2462] The server exports the simulation results as a dataset and sends it to the terminal.

[2463] I...

Claims

1. A means of collecting the careers, past decisions, statements, and related literature of executives and generating them as a unified dataset, A method for training a machine learning model based on preprocessed data and generating a digital twin of executives, A means of analyzing data and scenarios related to new management decisions and performing simulations, A system that includes means for visually displaying the results of a simulation to the user.

2. The system according to claim 1 for generating individual digital twin models of executives.

3. The system according to claim 1, comprising a terminal for inputting data related to new business decisions, and a server that analyzes the data, generates prediction results, and displays them.

Images