system

Abstract

We provide the system. [Solution] A means of collecting data on the past behavior of managers within a company, Means for preprocessing the collected data, A means for training a generative artificial intelligence model using preprocessed data, Means of receiving inquiries from users, A means of processing received inquiries using a generative artificial intelligence model, Means for providing the processing results to the user, A system that includes this.

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, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, 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] After the retirement of a corporate manager, especially the chief executive officer (CEO), there is a need for a method to utilize their rich knowledge and experience within the organization without loss. In conventional technologies, there is a problem that means for utilizing the opinions and decision-making processes of retired CEOs are limited, making it difficult for organizations to effectively inherit such valuable resources. Also, since means for new managers to refer to past effective decision-making and strategies are insufficient, there is a risk that the same strategies cannot be replicated or that responses to new challenges are inconsistent.

Means for Solving the Problems

[0005] This invention provides a system for constructing a virtual CEO by collecting past behavioral data of corporate managers, particularly CEOs, and training a generative artificial intelligence model using pre-processed data. This system includes means for receiving user inquiries, processing them with the generative AI model, and providing the processing results to the user. This configuration enables the effective sharing and utilization of the knowledge and experience of retired managers within the organization. Furthermore, by utilizing natural language processing technology, users can ask questions to the virtual CEO in a natural conversational format and obtain specific and accurate answers. This ensures consistency in the decision-making process within the organization and allows for the referencing of past effective strategies and decisions.

[0006] A "company" is a legal entity or individual that has an organized structure and engages in economic activities to produce goods or provide services.

[0007] A "manager" is a person in a position to direct and supervise the operation and management of a company or organization.

[0008] A "Chief Executive Officer (CEO)" is a person who, as the highest-ranking officer of a company, is responsible for overseeing the management and strategy of the entire company.

[0009] "Behavioral data" refers to recorded information about specific actions and decisions taken by administrators.

[0010] "Preprocessing" refers to the process of converting collected data into a format that can be used by the AI ​​model for learning, and includes tasks such as data cleaning and formatting adjustments.

[0011] A "generative artificial intelligence model" is an artificial intelligence model that has the ability to understand and generate human language.

[0012] A "user" is an individual or group that operates a system and provides or receives information.

[0013] An "inquiry" is a question or consultation that a user makes to a system in order to seek information.

[0014] "Natural language processing technology" refers to computer technology used to understand and generate human language, and includes text analysis and generation.

[0015] A "virtual CEO" is a model that virtually replicates the knowledge and experience of a CEO, constructed using a generative artificial intelligence model.

[0016] "Processing result" refers to the answer or information generated by a generative artificial intelligence model in response to an inquiry. [Brief explanation of the drawing]

[0017] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This 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] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This 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] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This 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] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Embodiment 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 Embodiment 2 when the 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 the emotion engine is combined.

Modes for Carrying Out the Invention

[0018] 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.

[0019] First, the terms used in the following description will be explained.

[0020] In the following embodiments, a numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple 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.

[0021] 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.

[0022] 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.

[0023] 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).

[0024] 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."

[0025] [First Embodiment]

[0026] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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".

[0038] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[0039] System configuration and program processing

[0040] Data Acquisition and AI Model Training

[0041] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[0042] User inquiry processing

[0043] Users input business-related questions and inquiries from their terminals to the server. For example, specific questions might include, "Please advise on the timing of the market launch of a new product." The questions sent from the terminals are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate an answer.

[0044] Generating and providing answers

[0045] The answers generated by the AI ​​model are validated by the server. This validation includes checking the appropriateness and grammar of the answers. The generated answers are adjusted as needed. Finally, the server sends the finalized answer to the device. The device then displays the received answer to the user. For example, if the user asks, "Give me some advice on the market strategy for a new product," the device might display specific advice such as, "We believe the optimal time to launch a new product is six months before competitors plan to release theirs."

[0046] Specific examples

[0047] Data Acquisition and Preprocessing

[0048] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[0049] Training an AI model

[0050] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as TENSORFLOW® and PyTorch. During the training process, the model learns from past speeches and decision-making examples to enable it to reproduce the CEO's thought patterns.

[0051] User inquiries and response generation

[0052] The user inputs "Please give me your opinion on future market strategies" from their device. The device sends this input to the server in JSON format. The server receives this question, extracts key keywords using NLP technology, and queries a generative AI model. The AI ​​model generates a response such as "Considering the actions of competitors, the optimal timing for market entry is six months ago."

[0053] Providing a response

[0054] The server verifies the generated response and, if appropriate, sends it directly to the terminal. The terminal displays the received response to the user, who then makes a decision based on the advice.

[0055] Thus, the system of the present invention provides a means to utilize the knowledge and experience of a company's past CEOs and to reflect their valuable opinions even after their retirement.

[0056] The following describes the processing flow.

[0057] Step 1:

[0058] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also uses APIs and scraping tools to collect data from the internet and other internal systems.

[0059] Step 2:

[0060] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. For example, the server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[0061] Step 3:

[0062] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[0063] Step 4:

[0064] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of launching a new product into the market."

[0065] Step 5:

[0066] The device sends user questions to the server in real time. During transmission, the question data is structured using JSON format, and the data is sent to the server using an HTTP request.

[0067] Step 6:

[0068] The server analyzes the questions received from users and classifies them into appropriate themes and categories. During this process, natural language processing (NLP) techniques are used to extract the main keywords of the questions and perform semantic analysis.

[0069] Step 7:

[0070] The server inputs a question into a generative AI model based on the analysis results, and generates an answer. The AI ​​model uses its past learning to generate a specific answer. For example, it might generate an answer such as, "The optimal time to launch a new product into the market is six months before competitors plan to release theirs."

[0071] Step 8:

[0072] The server validates the generated response. The validation process checks the appropriateness and grammar of the response and adjusts the content of the response as needed.

[0073] Step 9:

[0074] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[0075] Step 10:

[0076] The device displays the received responses to the user. The user reviews specific advice and opinions through the device's interface and makes decisions based on them. For example, they might adopt a strategy such as "setting the market launch date six months before the competitor's planned release."

[0077] Thus, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience through a series of processing steps.

[0078] (Example 1)

[0079] 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."

[0080] There is a need to effectively utilize the knowledge and experience of corporate managers, especially CEOs, even after their retirement, and to reflect them in current business strategies and decision-making. However, conventional methods are limited to mere data collection and analysis, and do not lead to actual decision-making support or the provision of concrete advice. To solve this problem and realize more sophisticated decision-making support, a new system utilizing generative AI models is necessary.

[0081] 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.

[0082] In this invention, the server includes means for collecting behavioral data of past corporate managers; means for preprocessing the collected data by standardizing it into a text format, deleting unnecessary information, and supplementing missing data; means for training a generative AI model using the preprocessed data; means for sending user inquiries to the server via a terminal; means for analyzing the received inquiries, extracting key keywords, inputting them into the generative AI model, and generating answers; means for verifying the generated answers on the server, checking grammar and content appropriateness, and making adjustments as necessary; and means for sending the final answers from the server to the terminal and displaying them to the user. This makes it possible to utilize the knowledge and experience of past corporate managers and reflect their valuable opinions in current corporate management even after their retirement.

[0083] "Data on past management behavior within a company" refers to information such as presentations, speeches, meeting notes, and decision-making records made by past managers within a company, particularly the Chief Executive Officer (CEO).

[0084] "Preprocessing to unify into text format, remove unnecessary information, and fill in missing data" refers to a series of processes that convert collected data into a unified text format, remove unnecessary tags and metadata, and fill in incomplete data.

[0085] "Training a generative AI model" refers to the process of training a generative artificial intelligence model using pre-processed data to learn the CEO's language style and decision-making patterns.

[0086] "Sending user inquiries to the server via the terminal" refers to a series of steps in which questions and inquiries entered by the user using their terminal are sent to the server in a data format such as JSON.

[0087] "Analyzing received inquiries, extracting key keywords, inputting them into a generative AI model, and generating answers" refers to the process by which a server analyzes inquiries received from users using natural language processing technology, extracts important keywords, inputs them into a generative AI model, and generates appropriate answers.

[0088] "Validating the generated responses on the server, checking grammar and content appropriateness, and making adjustments as needed" refers to the process of checking the accuracy and appropriateness of the grammar and content of responses generated by generative AI models, and making manual adjustments as necessary.

[0089] "Sending the final response from the server to the terminal and displaying it to the user" refers to a series of steps in which the server sends the verified and adjusted response to the terminal, and the terminal displays the received response to the user.

[0090] This invention is a system for collecting past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructing a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[0091] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. The hardware used includes internal servers and cloud servers. The collected data is standardized into text format using scripting languages ​​such as Python and Perl, and pre-processed by removing unnecessary information and filling in missing data.

[0092] The preprocessed data is fed into a generative artificial intelligence model (e.g., TensorFlow or PyTorch). This allows the model to learn the language styles and decision-making patterns of past CEOs. For example, presentation files from the past five years are extracted from the company's presentation database, unnecessary information is removed using regular expressions, and the data is standardized into text format.

[0093] A user sends a business-related question from their device to the server. For example, they might enter a prompt such as, "Please advise me on our market strategy for the next quarter." The server analyzes the received question using natural language processing technology and extracts key keywords. Based on these keywords, a generative AI model generates a prompt and then an answer.

[0094] The generated responses are validated on the server to check grammar and the appropriateness of the content. If there are any inappropriate parts, the server manually adjusts them. For example, if the AI ​​model responds, "The market launch timing is next year," the server verifies whether that information is accurate. The final confirmed response is sent to the terminal and displayed to the user.

[0095] The responses obtained from users via their devices can be used to inform business strategies and decision-making. For example, if a user asks, "Please give me your opinion on future market strategies," the device will display specific advice such as, "Considering the actions of competitors, six months prior to market launch would be appropriate."

[0096] This system makes it possible to leverage the knowledge and experience of past company managers and incorporate their valuable opinions into current company operations even after they have retired.

[0097] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0098] Step 1:

[0099] Data collection

[0100] The server collects past administrator activity data from internal corporate databases and publicly available documents. This primarily includes presentation files, meeting notes, and decision-making records. For example, the server periodically scans presentation files stored in the company's cloud storage for the past five years to collect new data. The input is the corporate database, and the output is the collected activity data.

[0101] Step 2:

[0102] Data preprocessing

[0103] The server collects data, standardizes it into text format, removes unnecessary information using regular expressions, and fills in missing data. As a specific example, it removes HTML tags and metadata from collected presentation data and converts it to text format. The input is collected behavioral data, and the output is pre-processed text data.

[0104] Step 3:

[0105] Training an AI model

[0106] The server uses pre-processed data to train a generative AI model. Here, frameworks such as TensorFlow and PyTorch are used to learn the administrator's language style and decision-making patterns. Specifically, the server uses a GPU to process a large dataset at high speed and train the model. The input is pre-processed text data, and the output is the trained generative AI model.

[0107] Step 4:

[0108] User inquiry processing

[0109] A user sends a business-related question to the server from their device. For example, they might enter the prompt, "Please advise on our market strategy for the next quarter." The input is the user's question text, and the output is this text converted into JSON format.

[0110] Step 5:

[0111] Question analysis

[0112] The server analyzes the received question using natural language processing techniques and extracts key keywords. As a specific example, the server extracts the keywords "market strategy" and "next quarter." The input is question data in JSON format, and the output is the extracted keywords.

[0113] Step 6:

[0114] Answer generation

[0115] The server generates prompt sentences for the AI ​​model based on the extracted keywords, and then inputs them into the model to generate a response. For example, if the keywords "market strategy, next quarter" are input into the model, it will generate the response "We recommend strengthening your competitive analysis." The input is the extracted keywords, and the output is the generated response text.

[0116] Step 7:

[0117] Verification of the answer

[0118] The server checks the grammar and appropriateness of the generated response and makes adjustments as needed. For example, if there are errors or inappropriate parts in the generated response, it will be manually corrected. The input is the generated response text, and the output is the validated final response.

[0119] Step 8:

[0120] Providing a response

[0121] The server sends the verified response to the terminal and displays it to the user. For example, the response displayed on the terminal might be, "As part of our market strategy for the next quarter, please consider obtaining a patent to differentiate our product." The input is the verified final response text, and the output is the final response displayed to the user.

[0122] (Application Example 1)

[0123] 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."

[0124] Companies need to immediately leverage the knowledge and experience of past managers, especially CEOs, to support on-the-ground decision-making. However, there is a lack of effective means to utilize the knowledge and experience of retired CEOs. Furthermore, while optimization suggestions based on real-time data are essential for factory operations, a suitable system for this purpose does not exist. To solve these problems, a system is needed that models the knowledge of past CEOs and links it with actual operational data.

[0125] 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.

[0126] In this invention, the server includes means for collecting past behavioral data of a company's managers, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for collecting factory operation data in real time, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, and means for providing the processing results to the user. This enables real-time decision support and optimization suggestions based on the knowledge and experience of past managers.

[0127] "Behavioral data" refers to data that includes information on presentations, speeches, decision-making processes, and other activities that company managers have conducted in the past.

[0128] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[0129] A "generative artificial intelligence model" is an AI model that is trained using collected and pre-processed data, and it uses natural language processing techniques to generate answers to user questions.

[0130] "User inquiries" refer to questions or consultations made by individuals within a factory or company seeking advice on specific operations or market strategies.

[0131] "Methods for collecting data in real time" refers to a system that uses sensors and cameras within the factory to instantly transmit factory operation data to a server.

[0132] "Means for processing inquiries using generative artificial intelligence models" refers to processing methods for receiving inquiries from users, analyzing them using generative artificial intelligence models, and deriving appropriate answers.

[0133] "Means of providing to the user" refers to a system for verifying the answers and suggestions generated by generative artificial intelligence models and presenting them to the user in an appropriate format.

[0134] This invention is a system that collects behavioral data of past corporate managers (especially CEOs) and constructs a virtual CEO using a generative artificial intelligence model. A specific example of this system is shown below.

[0135] Data Acquisition and Preprocessing

[0136] The server collects data on past administrator presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data undergoes preprocessing, including string organization, removal of unnecessary data, and imputation of missing data. Python and the Pandas library are used for this preprocessing.

[0137] Training an AI model

[0138] The server uses a pre-processed dataset to build a generative artificial intelligence model. Training is performed using frameworks such as TensorFlow and PyTorch. This model learns from past speeches and decision-making patterns and is trained to reproduce the CEO's thought patterns.

[0139] Data acquisition equipment

[0140] Sensors and cameras are placed throughout the factory. Operational data is collected from these devices in real time and transmitted to a server.

[0141] User inquiries and response generation

[0142] Factory workers and managers can ask questions to the virtual CEO via tablet devices or voice input devices. For example, a specific question might be, "Please give us your opinion on future market strategies." This question is sent from the tablet device to the server in JSON format.

[0143] Model-based processing and response generation

[0144] As mentioned earlier, the server analyzes this query using NLP technology and inputs it into a generative artificial intelligence model. The model generates an appropriate response based on the query. For example, it might generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[0145] Providing a response

[0146] The server validates the generated response and, if appropriate, sends it to the worker's tablet device. The tablet device displays the received response to the user, who then makes a decision based on the advice.

[0147] Examples of prompt statements

[0148] The worker provides the virtual CEO with the following prompt:

[0149] The utilization rate of our factory's production lines is decreasing. Based on past data, please provide advice on how to improve this situation.

[0150] Based on this prompt, the generative artificial intelligence model generates the following response:

[0151] According to the analysis, machine A on line 1 has been inadequately maintained for the past two weeks. Increasing the maintenance frequency is expected to improve the operating rate by 20%.

[0152] In this way, the system of this invention can leverage the knowledge and experience of past managers and, in conjunction with real-time factory data, provide optimal decision-making support.

[0153] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0154] Step 1: Data Collection

[0155] The server collects data on past administrator behavior from internal corporate databases and publicly available documents. This data includes records of presentations, speeches, and decision-making. The server stores the collected data in string format.

[0156] Input: Corporate databases, publicly available documents

[0157] Output: Behavioral data in string format

[0158] Step 2: Data Preprocessing

[0159] The server preprocesses the collected behavioral data. This preprocessing includes removing unnecessary data, imputing missing data, and standardizing the text format. This is done using Python and the Pandas library.

[0160] Input: Behavioral data in string format

[0161] Output: Preprocessed text data

[0162] Step 3: Training the AI ​​model

[0163] The server uses pre-processed data to train a generative artificial intelligence model. Using TensorFlow or PyTorch, the model learns the administrator's thought patterns and decision-making patterns.

[0164] Input: Preprocessed text data

[0165] Output: Trained generative AI model

[0166] Step 4: Real-time data collection

[0167] Sensors and cameras within the factory collect operational data in real time and transmit it to a server. This data includes machine operating status, temperature, humidity, and other information.

[0168] Input: Sensors and cameras within the factory

[0169] Output: Real-time operation data

[0170] Step 5: User Inquiry

[0171] Users can ask questions to the virtual CEO using tablet devices or voice input devices. For example, specific questions could include, "What are your thoughts on future market strategies?"

[0172] Input: User's question (e.g., "Please give us your opinion on future market strategies")

[0173] Output: Text data of the question

[0174] Step 6: Analyze the inquiry

[0175] The server analyzes the user's question and extracts key keywords using NLP techniques. These keywords are then input into a generative artificial intelligence model.

[0176] Input: Text data of the question

[0177] Output: Keyword extraction results using NLP technology

[0178] Step 7: Answer generation using AI model

[0179] Generative artificial intelligence models generate appropriate answers based on analyzed user questions. For example, they can generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[0180] Input: Keyword extraction results

[0181] Output: Text data of the response

[0182] Step 8: Verification of the answer

[0183] The server grammatically checks the generated responses and verifies their content validity. Once verification is complete, it selects the appropriate response and makes it the final answer.

[0184] Input: Text data of the response

[0185] Output: Verified text data

[0186] Step 9: Provide your response

[0187] The server sends verified answers to the user's tablet device. The device displays the received answers to the user. The user makes a decision based on the displayed advice.

[0188] Input: Verified text data

[0189] Output: Display of answers on a tablet device

[0190] In this way, it becomes possible to leverage the knowledge and experience of past managers and integrate it with real-time factory data to provide optimal decision-making support.

[0191] 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.

[0192] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[0193] System configuration and program processing

[0194] Data Acquisition and AI Model Training

[0195] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[0196] User inquiry processing and sentiment recognition

[0197] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of the market launch of a new product." The terminal has an emotion engine built in, which analyzes not only the user's input but also their emotional state at the time of input. The emotion engine detects the user's emotional state using voice analysis and facial recognition technology and sends the analysis results to the server.

[0198] Query and sentiment data processing

[0199] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate answers. The generative AI model generates answers while also considering the user's emotional state.

[0200] Generating and providing answers

[0201] The AI ​​model's generated responses are validated by the server. This validation includes checking the appropriateness and grammar of the responses. The generated responses are adjusted as needed. Finally, the server sends the finalized response to the device. The device then displays the received response to the user. For example, if the user is nervous, the response might be delivered in a softer tone, such as, "We believe that six months before your competitors' release dates is the appropriate time to launch your new product, but please let us know if you have any further questions."

[0202] Specific examples

[0203] Data Acquisition and Preprocessing

[0204] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[0205] Training an AI model

[0206] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as Tensorflow and PyTorch. During the training process, the model learns from past CEO speeches and decision-making examples to enable it to reproduce the thought patterns of CEOs.

[0207] User sentiment recognition and inquiries

[0208] The user types "Please give me your opinion on future market strategies" from their device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone, detecting their emotional state, such as whether they are tense or relaxed. The query data, including the analysis results, is sent to the server in JSON format.

[0209] Analysis and response generation on the server

[0210] Based on the data received by the server, it queries a generative AI model. The AI ​​model, taking into account the user's emotional state, generates a response such as, "Considering the actions of competitors, the optimal timing for market launch is six months ago."

[0211] Providing a response

[0212] The server verifies the generated response and, if appropriate, sends it directly to the device. The device displays the received response to the user, who then makes a decision based on the advice. For example, if the user is feeling anxious, the device might display a gentler message such as, "Set the market launch date to six months before your competitors' release schedule."

[0213] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby realizing a more human-like conversational experience.

[0214] The following describes the processing flow.

[0215] Step 1:

[0216] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also collects data from the internet and other internal systems using APIs and scraping tools.

[0217] Step 2:

[0218] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. The server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[0219] Step 3:

[0220] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[0221] Step 4:

[0222] Users input business inquiries and questions through the terminal's interface. For example, they might input, "Please advise me on the timing of the market launch for a new product."

[0223] Step 5:

[0224] The device's camera and microphone analyze the user's facial expressions and voice using an emotion engine to detect their emotional state. The device processes the analysis results in real time on the client side and sends them to the server along with the question data.

[0225] Step 6:

[0226] The device sends the user's questions and sentiment data to the server in JSON format. This transmission is done using an HTTP request.

[0227] Step 7:

[0228] The server analyzes the questions and sentiment data received from the user. Natural language processing (NLP) techniques are used to analyze the question content, extract key keywords, and classify them into appropriate themes and categories.

[0229] Step 8:

[0230] The server inputs questions and sentiment data into a generative AI model and instructs it to generate answers. The AI ​​model generates specific and contextually relevant answers while considering the user's emotional state. For example, it might generate an answer such as, "The optimal timing for market launch is six months before your competitors' release plans."

[0231] Step 9:

[0232] The server validates the generated response. This validation includes checking the appropriateness and grammar of the response. The server adjusts the content of the response as needed.

[0233] Step 10:

[0234] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[0235] Step 11:

[0236] The device displays the received response to the user. The emotion engine takes into account the user's emotional state, which it has detected in advance, and displays the response in an appropriate tone and wording. For example, if the user is nervous, the device will display in a gentle tone, "We recommend setting your market launch date six months before your competitors' release dates."

[0237] In this way, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience, and makes the conversation more human-like by taking the user's emotions into consideration.

[0238] (Example 2)

[0239] 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".

[0240] In modern business management, rapid and appropriate decision-making is essential. In particular, decision-making processes that leverage the CEO's past knowledge and experience are highly valuable. However, a system that integrates data collection, analysis, and response generation that considers user emotions does not exist. Furthermore, conventional artificial intelligence systems have struggled to generate responses that take user emotional states into account. This limits the user experience, highlighting the need for a system that can provide appropriate decision-making support.

[0241] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past behavioral data of the company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving user inquiries via a terminal, means for analyzing the received inquiries and classifying them into appropriate themes or categories, means including an emotion engine that analyzes the user's emotional state via the terminal at the time of inquiry, means for inputting the analyzed emotional state and inquiry content into the generative artificial intelligence model and processing it, means for verifying and adjusting the processing results, and means for providing the processing results to the user via the terminal. This makes it possible to provide a function as a virtual advisor that leverages the company's past knowledge and experience, and also enables the generation of detailed responses according to the user's emotional state.

[0242] A "company" is a legal entity or business that conducts economic activities as an organization.

[0243] A "manager" is a person who is responsible for effectively managing and operating operations, personnel, and resources within a company.

[0244] "Behavioral data" refers to information about work-related actions, such as managers' decision-making, statements made in meetings, and official documents.

[0245] "Preprocessing" refers to processes performed after data collection, such as deleting unnecessary information and standardizing data formats.

[0246] A "generative artificial intelligence model" is an artificial intelligence technology that learns from past data and generates responses in natural language for new data.

[0247] A "user" is an individual or organization that uses this system to receive information or support.

[0248] "Inquiry" refers to questions or inquiries entered by users into the system.

[0249] A "terminal" is a hardware device, such as a computer or smartphone, used to access and operate a system.

[0250] An "emotion engine" is a software technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state.

[0251] "Verification" refers to the process of checking whether the generated response is appropriate and grammatically correct.

[0252] "Providing" refers to the act of a server displaying or transmitting a generated response to a user.

[0253] This invention relates to a system that collects behavioral data from past corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[0254] Specific examples

[0255] Data Acquisition and Preprocessing

[0256] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data is standardized into a text format. As a preprocessing step, regular expressions are used to remove unnecessary information and standardize the data format.

[0257] Specific example: A server extracts CEO speeches, meeting notes, and decision-making records from the company's internal database over the past five years. After collection, it uses regular expressions to remove unnecessary metadata.

[0258] Training an AI model

[0259] The server uses frameworks such as TensorFlow and PyTorch to build a generative artificial intelligence model using a preprocessed dataset. By training this model with the behavioral patterns and language styles of past CEOs, a virtual CEO is created.

[0260] Specific example: The server defines a neural network using TensorFlow and trains it for 100 epochs using historical datasets.

[0261] User inquiry processing and sentiment recognition

[0262] Users input specific business inquiries and questions through the terminal's interface. The terminal is equipped with an emotion engine, which analyzes not only the user's input but also their emotional state at the time of input.

[0263] Specific example: The user types "Please give me your opinion on future market strategies" into the device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone to determine whether the user is tense or relaxed.

[0264] Query and sentiment data processing

[0265] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions and classifies them into appropriate themes and categories. It then inputs this data into a generative artificial intelligence model to generate responses that take the user's emotional state into account.

[0266] Specific example: The terminal sends the message "Please advise me on the timing of market launch" and emotion data indicating "tension" which are received by the server. The server categorizes the question under the theme of "market launch timing" and inputs it into a generative artificial intelligence model.

[0267] Generating and providing answers

[0268] The responses generated by the generative artificial intelligence model are validated by the server. This validation includes checking the appropriateness and grammar of the responses, and adjustments are made as needed. Finally, the server sends the finalized response to the terminal. The terminal then displays the received response to the user.

[0269] Specific example: The AI ​​model generates the response, "The optimal timing for market launch is six months in advance," which is validated by the server. After checking for appropriateness and grammar, the final response, "The appropriate time for market launch is six months before the competitor's planned release," is sent to the terminal. The terminal then displays this response to the user.

[0270] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby achieving a more human-like interaction.

[0271] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0272] Step 1: Data Collection

[0273] The server collects data on past CEO presentations, speeches, and decision-making from internal company databases and publicly available documents. Input is from internal company databases and online documents, from which data such as speeches, meeting notes, and decision-making records are extracted. The output is the collected, unprocessed behavioral data.

[0274] Step 2: Data preprocessing

[0275] The server converts the collected data into a unified text format and removes unnecessary information. The input is the raw behavioral data collected in step 1. The server uses regular expressions to remove unnecessary metadata and unify the data format into a consistent format. The output is the pre-processed text data.

[0276] Step 3: Training the AI ​​model

[0277] The server trains a generative artificial intelligence model using a preprocessed dataset. The input is the preprocessed text data obtained in step 2. Specifically, the model is built using TensorFlow or PyTorch and trained using the dataset. The output is the trained generative artificial intelligence model.

[0278] Step 4: Receiving user inquiry

[0279] The user enters specific business inquiries or questions through the terminal's interface. The input is a text-based question (prompt) entered by the user. The terminal receives this input and proceeds to the next step. The output is the user's inquiry text.

[0280] Step 5: Analysis of emotional state

[0281] The emotion engine analyzes the user's emotional state through the device's camera and microphone. The input is the user's facial expressions and tone of voice. The device uses facial recognition and voice analysis technologies to analyze the user's emotional state and extract emotional characteristics such as tension and relaxation. The output is the analyzed user's emotional state data.

[0282] Step 6: Data transmission

[0283] The terminal sends the user's inquiry and emotion data to the server. The input is the user's inquiry text obtained in Step 4 and the emotion state data obtained in Step 5. The terminal converts these data into JSON format and sends them to the server. The output is the JSON-formatted data sent to the server.

[0284] Step 7: Analysis and Categorization of Questions

[0285] Based on the data received by the server, the question is analyzed and classified into appropriate themes or categories. The input is the JSON-formatted data received from the terminal in Step 6. The server analyzes the question using natural language processing technology and classifies it into themes such as "market launch timing" and "competitive analysis". The output is the classified question data.

[0286] Step 8: Processing of Questions and Emotion Data

[0287] The server inputs the question and emotion state data into a generative artificial intelligence model and instructs it to generate an answer. The input is the question data classified in Step 7 and the emotion state data received in Step 6. The generative artificial intelligence model generates an answer based on these data. The output is the answer generated by the AI model.

[0288] Step 9: Verification and Adjustment of Answers

[0289] The server verifies the generated answer and adjusts it if necessary. The input is the answer generated in Step 8. The server checks the appropriateness and grammar of the answer and adjusts the content if necessary. The output is the verified and adjusted answer.

[0290] Step 10: Provision of Answers

[0291] The server provides the answer to the user via the terminal. The input is the verified and adjusted answer in Step 9. The server sends the finalized answer to the terminal, and the terminal displays the received answer to the user. The output is the answer displayed to the user.

[0292] (Application Example 2)

[0293] 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".

[0294] Traditionally, artificial intelligence models generated using behavioral data from corporate administrators could produce appropriate responses to inquiries, but they could not provide responses that took into account the user's emotional state. As a result, users, especially those experiencing tension or stress, were unable to receive appropriate advice, leading to a decline in the quality of decision-making.

[0295] 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.

[0296] In this invention, the server includes means for collecting past behavioral data of a company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, means for providing the processing results to the user, means for detecting the user's emotional state using an emotion engine, and means for reflecting the user's emotional state in the response generation of the generative artificial intelligence model. This makes it possible to provide responses that take the user's emotional state into consideration.

[0297] "Data on past management behavior within a company" refers to information including records of presentations, speeches, and decision-making activities previously conducted by managers within the company.

[0298] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[0299] The "generative AI model" is a type of artificial intelligence that is trained using collected and preprocessed data to reproduce the past behavior patterns and decision-making of managers.

[0300] The "emotion engine" is a device or software that includes technologies for analyzing facial expressions and voices to detect the emotional state of the user.

[0301] An "inquiry" is an act in which a user inputs specific questions or consultations regarding business.

[0302] The "processing result" is the content of the answer generated based on the generative AI model and the detected emotional state.

[0303] "Reflection" is an act in which the generative AI model adjusts its response considering the detected emotional state of the user.

[0304] This invention is a system that collects the past management behavior data of an enterprise and constructs a virtual manager using a generative AI model. Furthermore, by combining an emotion engine that recognizes the emotional state of the user, a more human-like response can be provided. Specific embodiments of this system will be described below.

[0305] Data Collection and Preprocessing

[0306] The server collects data on past managers' presentations, speeches, and decision-making from the enterprise's database. This data undergoes preprocessing such as unifying it into a text format using regular expressions, deleting unnecessary information, and complementing missing data. The preprocessed data is used as training data for the generative AI model.

[0307] Training of the Generative AI Model

[0308] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow and PyTorch. This training process analyzes past administrator behavior data to learn patterns of speech and decision-making styles.

[0309] Inquiry and Sentiment Recognition

[0310] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. These devices are equipped with an emotion engine that uses the camera and microphone to analyze the user's facial expressions and tone of voice, detecting their emotional state. For example, when a user inputs "Please tell me how to resolve the bottleneck in the production line," the emotion engine analyzes whether the user is tense or relaxed.

[0311] Query and sentiment data processing

[0312] The user's inquiry and emotional data are sent to the server in JSON format. The server analyzes the received data and inputs it into a generative artificial intelligence model. This model generates responses while taking the user's emotional state into consideration. For example, if the user is nervous, the response will be provided in a gentle tone.

[0313] Generating and providing answers

[0314] The AI ​​model's generated responses are validated by the server, and if appropriate, are sent directly to the user's device. The device then displays the received responses to the user. This entire process allows the user to receive appropriate advice from a virtual administrator. For example, specific responses such as, "To eliminate bottlenecks in the production line, it would be effective to review the worker shift schedule and utilize production management software," are provided.

[0315] Specific examples and prompt statements

[0316] As a concrete example, production reports from the past 10 years can be extracted from a factory's project logs. The user can then ask questions such as, "Please tell me specific ways to improve production efficiency," and the system will recognize their facial expressions using a camera and send the results to a server, providing appropriate advice. This allows the user to take concrete measures to improve productivity.

[0317] Example of a prompt

[0318] "Please tell me about specific measures to improve production efficiency."

[0319] "I'd like some advice on the timing of market launch."

[0320] "How can we optimize claims processing for a new production line?"

[0321] This invention makes it possible to provide appropriate decision-making support that takes into account the user's emotional state while utilizing the knowledge and experience of past administrators.

[0322] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0323] Step 1:

[0324] The server collects data from the company's database regarding past administrator presentations, speeches, and decisions. This involves a data collection process that executes database queries to extract the necessary information and convert it into an appropriate format. The input is the company's database, and the output is the collected raw data.

[0325] Step 2:

[0326] The server preprocesses the collected data. In this step, unnecessary information is removed using regular expressions, missing data is imputed, and the data is standardized to text format. This allows generative artificial intelligence models to efficiently learn from the data. The input is the collected raw data, and the output is the preprocessed clean data.

[0327] Step 3:

[0328] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow or PyTorch. During this process, the data is used as a training set for the model, learning the administrator's behavioral patterns and decision-making processes. The input is pre-processed clean data, and the output is the trained AI model.

[0329] Step 4:

[0330] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. The user's input is presented as natural language prompts. The input is a text query from the user, and the output is data sent from the device to the server.

[0331] Step 5:

[0332] The emotion engine built into the device uses the camera and microphone to analyze the user's facial expressions and tone of voice to detect their emotional state. This emotion data is sent to the server in JSON format. The input is the user's voice and video data, and the output is the analyzed emotion data.

[0333] Step 6:

[0334] The server analyzes the received question content and sentiment data, and inputs it into a generative artificial intelligence model. The model considers the user's emotional state and generates an appropriate response. The input is the received question content and sentiment data, and the output is the response generated by the model.

[0335] Step 7:

[0336] The server validates the generated response and makes corrections if necessary. This validation includes checking the appropriateness and grammar of the response. The input is the generated response, and the output is the validated or corrected response.

[0337] Step 8:

[0338] The server sends the final answer to the terminal. The terminal displays the received answer to the user. This sequence of actions allows the user to receive appropriate advice from the virtual administrator. The input is the verified or corrected answer, and the output is the final answer presented to the user.

[0339] 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.

[0340] 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.

[0341] 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.

[0342] [Second Embodiment]

[0343] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0344] 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.

[0345] 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).

[0346] 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.

[0347] 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.

[0348] 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).

[0349] 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.

[0350] 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.

[0351] 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.

[0352] 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.

[0353] 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.

[0354] 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".

[0355] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[0356] System configuration and program processing

[0357] Data Acquisition and AI Model Training

[0358] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[0359] User inquiry processing

[0360] Users input business-related questions and inquiries from their terminals to the server. For example, specific questions might include, "Please advise on the timing of the market launch of a new product." The questions sent from the terminals are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate an answer.

[0361] Generating and providing answers

[0362] The answers generated by the AI ​​model are validated by the server. This validation includes checking the appropriateness and grammar of the answers. The generated answers are adjusted as needed. Finally, the server sends the finalized answer to the device. The device then displays the received answer to the user. For example, if the user asks, "Give me some advice on the market strategy for a new product," the device might display specific advice such as, "We believe the optimal time to launch a new product is six months before competitors plan to release theirs."

[0363] Specific examples

[0364] Data Acquisition and Preprocessing

[0365] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[0366] Training an AI model

[0367] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as Tensorflow and PyTorch. During the training process, the model learns from past speeches and decision-making examples to enable it to reproduce the CEO's thought patterns.

[0368] User inquiries and response generation

[0369] The user inputs "Please give me your opinion on future market strategies" from their device. The device sends this input to the server in JSON format. The server receives this question, extracts key keywords using NLP technology, and queries a generative AI model. The AI ​​model generates a response such as "Considering the actions of competitors, the optimal timing for market entry is six months ago."

[0370] Providing a response

[0371] The server verifies the generated response and, if appropriate, sends it directly to the terminal. The terminal displays the received response to the user, who then makes a decision based on the advice.

[0372] Thus, the system of the present invention provides a means to utilize the knowledge and experience of a company's past CEOs and to reflect their valuable opinions even after their retirement.

[0373] The following describes the processing flow.

[0374] Step 1:

[0375] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also uses APIs and scraping tools to collect data from the internet and other internal systems.

[0376] Step 2:

[0377] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. For example, the server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[0378] Step 3:

[0379] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[0380] Step 4:

[0381] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of launching a new product into the market."

[0382] Step 5:

[0383] The device sends user questions to the server in real time. During transmission, the question data is structured using JSON format, and the data is sent to the server using an HTTP request.

[0384] Step 6:

[0385] The server analyzes the questions received from users and classifies them into appropriate themes and categories. During this process, natural language processing (NLP) techniques are used to extract the main keywords of the questions and perform semantic analysis.

[0386] Step 7:

[0387] The server inputs a question into a generative AI model based on the analysis results, and generates an answer. The AI ​​model uses its past learning to generate a specific answer. For example, it might generate an answer such as, "The optimal time to launch a new product into the market is six months before competitors plan to release theirs."

[0388] Step 8:

[0389] The server validates the generated response. The validation process checks the appropriateness and grammar of the response and adjusts the content of the response as needed.

[0390] Step 9:

[0391] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[0392] Step 10:

[0393] The device displays the received responses to the user. The user reviews specific advice and opinions through the device's interface and makes decisions based on them. For example, they might adopt a strategy such as "setting the market launch date six months before the competitor's planned release."

[0394] Thus, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience through a series of processing steps.

[0395] (Example 1)

[0396] 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."

[0397] There is a need to effectively utilize the knowledge and experience of corporate managers, especially CEOs, even after their retirement, and to reflect them in current business strategies and decision-making. However, conventional methods are limited to mere data collection and analysis, and do not lead to actual decision-making support or the provision of concrete advice. To solve this problem and realize more sophisticated decision-making support, a new system utilizing generative AI models is necessary.

[0398] 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.

[0399] In this invention, the server includes means for collecting behavioral data of past corporate managers; means for preprocessing the collected data by standardizing it into a text format, deleting unnecessary information, and supplementing missing data; means for training a generative AI model using the preprocessed data; means for sending user inquiries to the server via a terminal; means for analyzing the received inquiries, extracting key keywords, inputting them into the generative AI model, and generating answers; means for verifying the generated answers on the server, checking grammar and content appropriateness, and making adjustments as necessary; and means for sending the final answers from the server to the terminal and displaying them to the user. This makes it possible to utilize the knowledge and experience of past corporate managers and reflect their valuable opinions in current corporate management even after their retirement.

[0400] "Data on past management behavior within a company" refers to information such as presentations, speeches, meeting notes, and decision-making records made by past managers within a company, particularly the Chief Executive Officer (CEO).

[0401] "Preprocessing to unify into text format, remove unnecessary information, and fill in missing data" refers to a series of processes that convert collected data into a unified text format, remove unnecessary tags and metadata, and fill in incomplete data.

[0402] "Training a generative AI model" refers to the process of training a generative artificial intelligence model using pre-processed data to learn the CEO's language style and decision-making patterns.

[0403] "Sending user inquiries to the server via the terminal" refers to a series of steps in which questions and inquiries entered by the user using their terminal are sent to the server in a data format such as JSON.

[0404] "Analyzing received inquiries, extracting key keywords, inputting them into a generative AI model, and generating answers" refers to the process by which a server analyzes inquiries received from users using natural language processing technology, extracts important keywords, inputs them into a generative AI model, and generates appropriate answers.

[0405] "Validating the generated responses on the server, checking grammar and content appropriateness, and making adjustments as needed" refers to the process of checking the accuracy and appropriateness of the grammar and content of responses generated by generative AI models, and making manual adjustments as necessary.

[0406] "Sending the final response from the server to the terminal and displaying it to the user" refers to a series of steps in which the server sends the verified and adjusted response to the terminal, and the terminal displays the received response to the user.

[0407] This invention is a system for collecting past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructing a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[0408] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. The hardware used includes internal servers and cloud servers. The collected data is standardized into text format using scripting languages ​​such as Python and Perl, and pre-processed by removing unnecessary information and filling in missing data.

[0409] The preprocessed data is fed into a generative artificial intelligence model (e.g., TensorFlow or PyTorch). This allows the model to learn the language styles and decision-making patterns of past CEOs. For example, presentation files from the past five years are extracted from the company's presentation database, unnecessary information is removed using regular expressions, and the data is standardized into text format.

[0410] A user sends a business-related question from their device to the server. For example, they might enter a prompt such as, "Please advise me on our market strategy for the next quarter." The server analyzes the received question using natural language processing technology and extracts key keywords. Based on these keywords, a generative AI model generates a prompt and then an answer.

[0411] The generated responses are validated on the server to check grammar and the appropriateness of the content. If there are any inappropriate parts, the server manually adjusts them. For example, if the AI ​​model responds, "The market launch timing is next year," the server verifies whether that information is accurate. The final confirmed response is sent to the terminal and displayed to the user.

[0412] The responses obtained from users via their devices can be used to inform business strategies and decision-making. For example, if a user asks, "Please give me your opinion on future market strategies," the device will display specific advice such as, "Considering the actions of competitors, six months prior to market launch would be appropriate."

[0413] This system makes it possible to leverage the knowledge and experience of past company managers and incorporate their valuable opinions into current company operations even after they have retired.

[0414] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0415] Step 1:

[0416] Data collection

[0417] The server collects past administrator activity data from internal corporate databases and publicly available documents. This primarily includes presentation files, meeting notes, and decision-making records. For example, the server periodically scans presentation files stored in the company's cloud storage for the past five years to collect new data. The input is the corporate database, and the output is the collected activity data.

[0418] Step 2:

[0419] Data preprocessing

[0420] The server collects data, standardizes it into text format, removes unnecessary information using regular expressions, and fills in missing data. As a specific example, it removes HTML tags and metadata from collected presentation data and converts it to text format. The input is collected behavioral data, and the output is pre-processed text data.

[0421] Step 3:

[0422] Training an AI model

[0423] The server uses pre-processed data to train a generative AI model. Here, frameworks such as TensorFlow and PyTorch are used to learn the administrator's language style and decision-making patterns. Specifically, the server uses a GPU to process a large dataset at high speed and train the model. The input is pre-processed text data, and the output is the trained generative AI model.

[0424] Step 4:

[0425] User inquiry processing

[0426] A user sends a business-related question to the server from their device. For example, they might enter the prompt, "Please advise on our market strategy for the next quarter." The input is the user's question text, and the output is this text converted into JSON format.

[0427] Step 5:

[0428] Question analysis

[0429] The server analyzes the received question using natural language processing techniques and extracts key keywords. As a specific example, the server extracts the keywords "market strategy" and "next quarter." The input is question data in JSON format, and the output is the extracted keywords.

[0430] Step 6:

[0431] Answer generation

[0432] The server generates prompt sentences for the AI ​​model based on the extracted keywords, and then inputs them into the model to generate a response. For example, if the keywords "market strategy, next quarter" are input into the model, it will generate the response "We recommend strengthening your competitive analysis." The input is the extracted keywords, and the output is the generated response text.

[0433] Step 7:

[0434] Verification of the answer

[0435] The server checks the grammar and appropriateness of the generated response and makes adjustments as needed. For example, if there are errors or inappropriate parts in the generated response, it will be manually corrected. The input is the generated response text, and the output is the validated final response.

[0436] Step 8:

[0437] Providing a response

[0438] The server sends the verified response to the terminal and displays it to the user. For example, the response displayed on the terminal might be, "As part of our market strategy for the next quarter, please consider obtaining a patent to differentiate our product." The input is the verified final response text, and the output is the final response displayed to the user.

[0439] (Application Example 1)

[0440] 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."

[0441] Companies need to immediately leverage the knowledge and experience of past managers, especially CEOs, to support on-the-ground decision-making. However, there is a lack of effective means to utilize the knowledge and experience of retired CEOs. Furthermore, while optimization suggestions based on real-time data are essential for factory operations, a suitable system for this purpose does not exist. To solve these problems, a system is needed that models the knowledge of past CEOs and links it with actual operational data.

[0442] 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.

[0443] In this invention, the server includes means for collecting past behavioral data of a company's managers, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for collecting factory operation data in real time, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, and means for providing the processing results to the user. This enables real-time decision support and optimization suggestions based on the knowledge and experience of past managers.

[0444] "Behavioral data" refers to data that includes information on presentations, speeches, decision-making processes, and other activities that company managers have conducted in the past.

[0445] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[0446] A "generative artificial intelligence model" is an AI model that is trained using collected and pre-processed data, and it uses natural language processing techniques to generate answers to user questions.

[0447] "User inquiries" refer to questions or consultations made by individuals within a factory or company seeking advice on specific operations or market strategies.

[0448] "Methods for collecting data in real time" refers to a system that uses sensors and cameras within the factory to instantly transmit factory operation data to a server.

[0449] "Means for processing inquiries using generative artificial intelligence models" refers to processing methods for receiving inquiries from users, analyzing them using generative artificial intelligence models, and deriving appropriate answers.

[0450] "Means of providing to the user" refers to a system for verifying the answers and suggestions generated by generative artificial intelligence models and presenting them to the user in an appropriate format.

[0451] This invention is a system that collects behavioral data of past corporate managers (especially CEOs) and constructs a virtual CEO using a generative artificial intelligence model. A specific example of this system is shown below.

[0452] Data Acquisition and Preprocessing

[0453] The server collects data on past administrator presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data undergoes preprocessing, including string organization, removal of unnecessary data, and imputation of missing data. Python and the Pandas library are used for this preprocessing.

[0454] Training an AI model

[0455] The server uses a pre-processed dataset to build a generative artificial intelligence model. Training is performed using frameworks such as TensorFlow and PyTorch. This model learns from past speeches and decision-making patterns and is trained to reproduce the CEO's thought patterns.

[0456] Data acquisition equipment

[0457] Sensors and cameras are placed throughout the factory. Operational data is collected from these devices in real time and transmitted to a server.

[0458] User inquiries and response generation

[0459] Factory workers and managers can ask questions to the virtual CEO via tablet devices or voice input devices. For example, a specific question might be, "Please give us your opinion on future market strategies." This question is sent from the tablet device to the server in JSON format.

[0460] Model-based processing and response generation

[0461] As mentioned earlier, the server analyzes this query using NLP technology and inputs it into a generative artificial intelligence model. The model generates an appropriate response based on the query. For example, it might generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[0462] Providing a response

[0463] The server validates the generated response and, if appropriate, sends it to the worker's tablet device. The tablet device displays the received response to the user, who then makes a decision based on the advice.

[0464] Examples of prompt statements

[0465] The worker provides the virtual CEO with the following prompt:

[0466] The utilization rate of our factory's production lines is decreasing. Based on past data, please provide advice on how to improve this situation.

[0467] Based on this prompt, the generative artificial intelligence model generates the following response:

[0468] According to the analysis, machine A on line 1 has been inadequately maintained for the past two weeks. Increasing the maintenance frequency is expected to improve the operating rate by 20%.

[0469] In this way, the system of this invention can leverage the knowledge and experience of past managers and, in conjunction with real-time factory data, provide optimal decision-making support.

[0470] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0471] Step 1: Data Collection

[0472] The server collects data on past administrator behavior from internal corporate databases and publicly available documents. This data includes records of presentations, speeches, and decision-making. The server stores the collected data in string format.

[0473] Input: Corporate databases, publicly available documents

[0474] Output: Behavioral data in string format

[0475] Step 2: Data Preprocessing

[0476] The server preprocesses the collected behavioral data. This preprocessing includes removing unnecessary data, imputing missing data, and standardizing the text format. This is done using Python and the Pandas library.

[0477] Input: Behavioral data in string format

[0478] Output: Preprocessed text data

[0479] Step 3: Training the AI ​​model

[0480] The server uses pre-processed data to train a generative artificial intelligence model. Using TensorFlow or PyTorch, the model learns the administrator's thought patterns and decision-making patterns.

[0481] Input: Preprocessed text data

[0482] Output: Trained generative AI model

[0483] Step 4: Real-time data collection

[0484] Sensors and cameras within the factory collect operational data in real time and transmit it to a server. This data includes machine operating status, temperature, humidity, and other information.

[0485] Input: Sensors and cameras within the factory

[0486] Output: Real-time operation data

[0487] Step 5: User Inquiry

[0488] Users can ask questions to the virtual CEO using tablet devices or voice input devices. For example, specific questions could include, "What are your thoughts on future market strategies?"

[0489] Input: User's question (e.g., "Please give us your opinion on future market strategies")

[0490] Output: Text data of the question

[0491] Step 6: Analyze the inquiry

[0492] The server analyzes the user's question and extracts key keywords using NLP techniques. These keywords are then input into a generative artificial intelligence model.

[0493] Input: Text data of the question

[0494] Output: Keyword extraction results using NLP technology

[0495] Step 7: Answer generation using AI model

[0496] Generative artificial intelligence models generate appropriate answers based on analyzed user questions. For example, they can generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[0497] Input: Keyword extraction results

[0498] Output: Text data of the response

[0499] Step 8: Verification of the answer

[0500] The server grammatically checks the generated responses and verifies their content validity. Once verification is complete, it selects the appropriate response and makes it the final answer.

[0501] Input: Text data of the response

[0502] Output: Verified text data

[0503] Step 9: Provide your response

[0504] The server sends verified answers to the user's tablet device. The device displays the received answers to the user. The user makes a decision based on the displayed advice.

[0505] Input: Verified text data

[0506] Output: Display of answers on a tablet device

[0507] In this way, it becomes possible to leverage the knowledge and experience of past managers and integrate it with real-time factory data to provide optimal decision-making support.

[0508] 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.

[0509] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[0510] System configuration and program processing

[0511] Data Acquisition and AI Model Training

[0512] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[0513] User inquiry processing and sentiment recognition

[0514] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of the market launch of a new product." The terminal has an emotion engine built in, which analyzes not only the user's input but also their emotional state at the time of input. The emotion engine detects the user's emotional state using voice analysis and facial recognition technology and sends the analysis results to the server.

[0515] Query and sentiment data processing

[0516] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate answers. The generative AI model generates answers while also considering the user's emotional state.

[0517] Generating and providing answers

[0518] The AI ​​model's generated responses are validated by the server. This validation includes checking the appropriateness and grammar of the responses. The generated responses are adjusted as needed. Finally, the server sends the finalized response to the device. The device then displays the received response to the user. For example, if the user is nervous, the response might be delivered in a softer tone, such as, "We believe that six months before your competitors' release dates is the appropriate time to launch your new product, but please let us know if you have any further questions."

[0519] Specific examples

[0520] Data Acquisition and Preprocessing

[0521] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[0522] Training an AI model

[0523] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as Tensorflow and PyTorch. During the training process, the model learns from past CEO speeches and decision-making examples to enable it to reproduce the thought patterns of CEOs.

[0524] User sentiment recognition and inquiries

[0525] The user types "Please give me your opinion on future market strategies" from their device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone, detecting their emotional state, such as whether they are tense or relaxed. The query data, including the analysis results, is sent to the server in JSON format.

[0526] Analysis and response generation on the server

[0527] Based on the data received by the server, it queries a generative AI model. The AI ​​model, taking into account the user's emotional state, generates a response such as, "Considering the actions of competitors, the optimal timing for market launch is six months ago."

[0528] Providing a response

[0529] The server verifies the generated response and, if appropriate, sends it directly to the device. The device displays the received response to the user, who then makes a decision based on the advice. For example, if the user is feeling anxious, the device might display a gentler message such as, "Set the market launch date to six months before your competitors' release schedule."

[0530] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby realizing a more human-like conversational experience.

[0531] The following describes the processing flow.

[0532] Step 1:

[0533] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also collects data from the internet and other internal systems using APIs and scraping tools.

[0534] Step 2:

[0535] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. The server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[0536] Step 3:

[0537] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[0538] Step 4:

[0539] Users input business inquiries and questions through the terminal's interface. For example, they might input, "Please advise me on the timing of the market launch for a new product."

[0540] Step 5:

[0541] The device's camera and microphone analyze the user's facial expressions and voice using an emotion engine to detect their emotional state. The device processes the analysis results in real time on the client side and sends them to the server along with the question data.

[0542] Step 6:

[0543] The device sends the user's questions and sentiment data to the server in JSON format. This transmission is done using an HTTP request.

[0544] Step 7:

[0545] The server analyzes the questions and sentiment data received from the user. Natural language processing (NLP) techniques are used to analyze the question content, extract key keywords, and classify them into appropriate themes and categories.

[0546] Step 8:

[0547] The server inputs questions and sentiment data into a generative AI model and instructs it to generate answers. The AI ​​model generates specific and contextually relevant answers while considering the user's emotional state. For example, it might generate an answer such as, "The optimal timing for market launch is six months before your competitors' release plans."

[0548] Step 9:

[0549] The server validates the generated response. This validation includes checking the appropriateness and grammar of the response. The server adjusts the content of the response as needed.

[0550] Step 10:

[0551] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[0552] Step 11:

[0553] The device displays the received response to the user. The emotion engine takes into account the user's emotional state, which it has detected in advance, and displays the response in an appropriate tone and wording. For example, if the user is nervous, the device will display in a gentle tone, "We recommend setting your market launch date six months before your competitors' release dates."

[0554] In this way, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience, and makes the conversation more human-like by taking the user's emotions into consideration.

[0555] (Example 2)

[0556] 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".

[0557] In modern business management, rapid and appropriate decision-making is essential. In particular, decision-making processes that leverage the CEO's past knowledge and experience are highly valuable. However, a system that integrates data collection, analysis, and response generation that considers user emotions does not exist. Furthermore, conventional artificial intelligence systems have struggled to generate responses that take user emotional states into account. This limits the user experience, highlighting the need for a system that can provide appropriate decision-making support.

[0558] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past behavioral data of the company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving user inquiries via a terminal, means for analyzing the received inquiries and classifying them into appropriate themes or categories, means including an emotion engine that analyzes the user's emotional state via the terminal at the time of inquiry, means for inputting the analyzed emotional state and inquiry content into the generative artificial intelligence model and processing it, means for verifying and adjusting the processing results, and means for providing the processing results to the user via the terminal. This makes it possible to provide a function as a virtual advisor that leverages the company's past knowledge and experience, and also enables the generation of detailed responses according to the user's emotional state.

[0559] A "company" is a legal entity or business that conducts economic activities as an organization.

[0560] A "manager" is a person who is responsible for effectively managing and operating operations, personnel, and resources within a company.

[0561] "Behavioral data" refers to information about work-related actions, such as managers' decision-making, statements made in meetings, and official documents.

[0562] "Preprocessing" refers to processes performed after data collection, such as deleting unnecessary information and standardizing data formats.

[0563] A "generative artificial intelligence model" is an artificial intelligence technology that learns from past data and generates responses in natural language for new data.

[0564] A "user" is an individual or organization that uses this system to receive information or support.

[0565] "Inquiry" refers to questions or inquiries entered by users into the system.

[0566] A "terminal" is a hardware device, such as a computer or smartphone, used to access and operate a system.

[0567] An "emotion engine" is a software technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state.

[0568] "Verification" refers to the process of checking whether the generated response is appropriate and grammatically correct.

[0569] "Providing" refers to the act of a server displaying or transmitting a generated response to a user.

[0570] This invention relates to a system that collects behavioral data from past corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[0571] Specific examples

[0572] Data Acquisition and Preprocessing

[0573] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data is standardized into a text format. As a preprocessing step, regular expressions are used to remove unnecessary information and standardize the data format.

[0574] Specific example: A server extracts CEO speeches, meeting notes, and decision-making records from the company's internal database over the past five years. After collection, it uses regular expressions to remove unnecessary metadata.

[0575] Training an AI model

[0576] The server uses frameworks such as TensorFlow and PyTorch to build a generative artificial intelligence model using a preprocessed dataset. By training this model with the behavioral patterns and language styles of past CEOs, a virtual CEO is created.

[0577] Specific example: The server defines a neural network using TensorFlow and trains it for 100 epochs using historical datasets.

[0578] User inquiry processing and sentiment recognition

[0579] Users input specific business inquiries and questions through the terminal's interface. The terminal is equipped with an emotion engine, which analyzes not only the user's input but also their emotional state at the time of input.

[0580] Specific example: The user types "Please give me your opinion on future market strategies" into the device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone to determine whether the user is tense or relaxed.

[0581] Query and sentiment data processing

[0582] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions and classifies them into appropriate themes and categories. It then inputs this data into a generative artificial intelligence model to generate responses that take the user's emotional state into account.

[0583] Specific example: The terminal sends the message "Please advise me on the timing of market launch" and emotion data indicating "tension" which are received by the server. The server categorizes the question under the theme of "market launch timing" and inputs it into a generative artificial intelligence model.

[0584] Generating and providing answers

[0585] The responses generated by the generative artificial intelligence model are validated by the server. This validation includes checking the appropriateness and grammar of the responses, and adjustments are made as needed. Finally, the server sends the finalized response to the terminal. The terminal then displays the received response to the user.

[0586] Specific example: The AI ​​model generates the response, "The optimal timing for market launch is six months in advance," which is validated by the server. After checking for appropriateness and grammar, the final response, "The appropriate time for market launch is six months before the competitor's planned release," is sent to the terminal. The terminal then displays this response to the user.

[0587] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby achieving a more human-like interaction.

[0588] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0589] Step 1: Data Collection

[0590] The server collects data on past CEO presentations, speeches, and decision-making from internal company databases and publicly available documents. Input is from internal company databases and online documents, from which data such as speeches, meeting notes, and decision-making records are extracted. The output is the collected, unprocessed behavioral data.

[0591] Step 2: Data preprocessing

[0592] The server converts the collected data into a unified text format and removes unnecessary information. The input is the raw behavioral data collected in step 1. The server uses regular expressions to remove unnecessary metadata and unify the data format into a consistent format. The output is the pre-processed text data.

[0593] Step 3: Training the AI ​​model

[0594] The server trains a generative artificial intelligence model using a preprocessed dataset. The input is the preprocessed text data obtained in step 2. Specifically, the model is built using TensorFlow or PyTorch and trained using the dataset. The output is the trained generative artificial intelligence model.

[0595] Step 4: Receiving user inquiry

[0596] The user enters specific business inquiries or questions through the terminal's interface. The input is a text-based question (prompt) entered by the user. The terminal receives this input and proceeds to the next step. The output is the user's inquiry text.

[0597] Step 5: Analysis of emotional state

[0598] The emotion engine analyzes the user's emotional state through the device's camera and microphone. The input is the user's facial expressions and tone of voice. The device uses facial recognition and voice analysis technologies to analyze the user's emotional state and extract emotional characteristics such as tension and relaxation. The output is the analyzed user's emotional state data.

[0599] Step 6: Data transmission

[0600] The terminal sends the user's inquiry and sentiment data to the server. The input is the user's inquiry text obtained in step 4 and the sentiment state data obtained in step 5. The terminal converts this data into JSON format and sends it to the server. The output is the JSON data sent to the server.

[0601] Step 7: Question Analysis and Categorization

[0602] The server analyzes the received data and classifies the questions into appropriate themes and categories. The input is the JSON data received from the terminal in step 6. The server uses natural language processing technology to analyze the questions and classify them into themes such as "timing of market entry" and "competitive analysis." The output is the classified question data.

[0603] Step 8: Processing Questions and Sentimental Data

[0604] The server inputs question and emotion state data into a generative artificial intelligence model and instructs it to generate an answer. The input consists of the question data classified in step 7 and the emotion state data received in step 6. The generative artificial intelligence model generates an answer based on this data. The output is the answer generated by the AI ​​model.

[0605] Step 9: Verify and adjust your response

[0606] The server validates the generated response and adjusts it as needed. The input is the response generated in step 8. The server checks the appropriateness and grammar of the response and adjusts the content if necessary. The output is the validated and adjusted response.

[0607] Step 10: Provide your response

[0608] The server provides the answer to the user via the terminal. The input is the answer that was verified and adjusted in step 9. The server sends the finalized answer to the terminal, and the terminal displays the received answer to the user. The output is the answer displayed to the user.

[0609] (Application Example 2)

[0610] 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."

[0611] Traditionally, artificial intelligence models generated using behavioral data from corporate administrators could produce appropriate responses to inquiries, but they could not provide responses that took into account the user's emotional state. As a result, users, especially those experiencing tension or stress, were unable to receive appropriate advice, leading to a decline in the quality of decision-making.

[0612] 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.

[0613] In this invention, the server includes means for collecting past behavioral data of a company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, means for providing the processing results to the user, means for detecting the user's emotional state using an emotion engine, and means for reflecting the user's emotional state in the response generation of the generative artificial intelligence model. This makes it possible to provide responses that take the user's emotional state into consideration.

[0614] "Data on past management behavior within a company" refers to information including records of presentations, speeches, and decision-making activities previously conducted by managers within the company.

[0615] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[0616] A "generative artificial intelligence model" is a type of artificial intelligence that is trained using collected and pre-processed data to reproduce past behavioral patterns and decision-making of administrators.

[0617] An "emotion engine" is a device or software that includes technology to analyze facial expressions and voice in order to detect the user's emotional state.

[0618] An "inquiry" is the act of a user entering specific questions or requests for advice regarding a business.

[0619] "Processing result" refers to the response content generated based on the generative artificial intelligence model and the detected emotional state.

[0620] "Reflection" refers to the process by which a generative artificial intelligence model adjusts its response in consideration of the detected emotional state of the user.

[0621] This invention is a system that collects past behavioral data of corporate managers and constructs a virtual manager using a generative artificial intelligence model. Furthermore, by combining it with an emotion engine that recognizes the user's emotional state, it can provide more human-like responses. A specific embodiment of this system is described below.

[0622] Data Acquisition and Preprocessing

[0623] The server collects data from the company's database regarding past administrator presentations, speeches, and decision-making. This data undergoes preprocessing, including standardization to text format using regular expressions, removal of unnecessary information, and imputation of missing data. The preprocessed data is then used as training data for generative artificial intelligence models.

[0624] Training of generative AI models

[0625] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow and PyTorch. This training process analyzes past administrator behavior data to learn patterns of speech and decision-making styles.

[0626] Inquiry and Sentiment Recognition

[0627] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. These devices are equipped with an emotion engine that uses the camera and microphone to analyze the user's facial expressions and tone of voice, detecting their emotional state. For example, when a user inputs "Please tell me how to resolve the bottleneck in the production line," the emotion engine analyzes whether the user is tense or relaxed.

[0628] Query and sentiment data processing

[0629] The user's inquiry and emotional data are sent to the server in JSON format. The server analyzes the received data and inputs it into a generative artificial intelligence model. This model generates responses while taking the user's emotional state into consideration. For example, if the user is nervous, the response will be provided in a gentle tone.

[0630] Generating and providing answers

[0631] The AI ​​model's generated responses are validated by the server, and if appropriate, are sent directly to the user's device. The device then displays the received responses to the user. This entire process allows the user to receive appropriate advice from a virtual administrator. For example, specific responses such as, "To eliminate bottlenecks in the production line, it would be effective to review the worker shift schedule and utilize production management software," are provided.

[0632] Specific examples and prompt statements

[0633] As a concrete example, production reports from the past 10 years can be extracted from a factory's project logs. The user can then ask questions such as, "Please tell me specific ways to improve production efficiency," and the system will recognize their facial expressions using a camera and send the results to a server, providing appropriate advice. This allows the user to take concrete measures to improve productivity.

[0634] Example of a prompt

[0635] "Please tell me about specific measures to improve production efficiency."

[0636] "I'd like some advice on the timing of market launch."

[0637] "How can we optimize claims processing for a new production line?"

[0638] This invention makes it possible to provide appropriate decision-making support that takes into account the user's emotional state while utilizing the knowledge and experience of past administrators.

[0639] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0640] Step 1:

[0641] The server collects data from the company's database regarding past administrator presentations, speeches, and decisions. This involves a data collection process that executes database queries to extract the necessary information and convert it into an appropriate format. The input is the company's database, and the output is the collected raw data.

[0642] Step 2:

[0643] The server preprocesses the collected data. In this step, unnecessary information is removed using regular expressions, missing data is imputed, and the data is standardized to text format. This allows generative artificial intelligence models to efficiently learn from the data. The input is the collected raw data, and the output is the preprocessed clean data.

[0644] Step 3:

[0645] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow or PyTorch. During this process, the data is used as a training set for the model, learning the administrator's behavioral patterns and decision-making processes. The input is pre-processed clean data, and the output is the trained AI model.

[0646] Step 4:

[0647] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. The user's input is presented as natural language prompts. The input is a text query from the user, and the output is data sent from the device to the server.

[0648] Step 5:

[0649] The emotion engine built into the device uses the camera and microphone to analyze the user's facial expressions and tone of voice to detect their emotional state. This emotion data is sent to the server in JSON format. The input is the user's voice and video data, and the output is the analyzed emotion data.

[0650] Step 6:

[0651] The server analyzes the received question content and sentiment data, and inputs it into a generative artificial intelligence model. The model considers the user's emotional state and generates an appropriate response. The input is the received question content and sentiment data, and the output is the response generated by the model.

[0652] Step 7:

[0653] The server validates the generated response and makes corrections if necessary. This validation includes checking the appropriateness and grammar of the response. The input is the generated response, and the output is the validated or corrected response.

[0654] Step 8:

[0655] The server sends the final answer to the terminal. The terminal displays the received answer to the user. This sequence of actions allows the user to receive appropriate advice from the virtual administrator. The input is the verified or corrected answer, and the output is the final answer presented to the user.

[0656] 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.

[0657] 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.

[0658] 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.

[0659] [Third Embodiment]

[0660] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0661] 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.

[0662] 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).

[0663] 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.

[0664] 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.

[0665] 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).

[0666] 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.

[0667] 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.

[0668] 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.

[0669] 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.

[0670] 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.

[0671] 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".

[0672] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[0673] System configuration and program processing

[0674] Data Acquisition and AI Model Training

[0675] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[0676] User inquiry processing

[0677] Users input business-related questions and inquiries from their terminals to the server. For example, specific questions might include, "Please advise on the timing of the market launch of a new product." The questions sent from the terminals are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate an answer.

[0678] Generating and providing answers

[0679] The answers generated by the AI ​​model are validated by the server. This validation includes checking the appropriateness and grammar of the answers. The generated answers are adjusted as needed. Finally, the server sends the finalized answer to the device. The device then displays the received answer to the user. For example, if the user asks, "Give me some advice on the market strategy for a new product," the device might display specific advice such as, "We believe the optimal time to launch a new product is six months before competitors plan to release theirs."

[0680] Specific examples

[0681] Data Acquisition and Preprocessing

[0682] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[0683] Training an AI model

[0684] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as Tensorflow and PyTorch. During the training process, the model learns from past speeches and decision-making examples to enable it to reproduce the CEO's thought patterns.

[0685] User inquiries and response generation

[0686] The user inputs "Please give me your opinion on future market strategies" from their device. The device sends this input to the server in JSON format. The server receives this question, extracts key keywords using NLP technology, and queries a generative AI model. The AI ​​model generates a response such as "Considering the actions of competitors, the optimal timing for market entry is six months ago."

[0687] Providing a response

[0688] The server verifies the generated response and, if appropriate, sends it directly to the terminal. The terminal displays the received response to the user, who then makes a decision based on the advice.

[0689] Thus, the system of the present invention provides a means to utilize the knowledge and experience of a company's past CEOs and to reflect their valuable opinions even after their retirement.

[0690] The following describes the processing flow.

[0691] Step 1:

[0692] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also uses APIs and scraping tools to collect data from the internet and other internal systems.

[0693] Step 2:

[0694] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. For example, the server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[0695] Step 3:

[0696] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[0697] Step 4:

[0698] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of launching a new product into the market."

[0699] Step 5:

[0700] The device sends user questions to the server in real time. During transmission, the question data is structured using JSON format, and the data is sent to the server using an HTTP request.

[0701] Step 6:

[0702] The server analyzes the questions received from users and classifies them into appropriate themes and categories. During this process, natural language processing (NLP) techniques are used to extract the main keywords of the questions and perform semantic analysis.

[0703] Step 7:

[0704] The server inputs a question into a generative AI model based on the analysis results, and generates an answer. The AI ​​model uses its past learning to generate a specific answer. For example, it might generate an answer such as, "The optimal time to launch a new product into the market is six months before competitors plan to release theirs."

[0705] Step 8:

[0706] The server validates the generated response. The validation process checks the appropriateness and grammar of the response and adjusts the content of the response as needed.

[0707] Step 9:

[0708] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[0709] Step 10:

[0710] The device displays the received responses to the user. The user reviews specific advice and opinions through the device's interface and makes decisions based on them. For example, they might adopt a strategy such as "setting the market launch date six months before the competitor's planned release."

[0711] Thus, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience through a series of processing steps.

[0712] (Example 1)

[0713] 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."

[0714] There is a need to effectively utilize the knowledge and experience of corporate managers, especially CEOs, even after their retirement, and to reflect them in current business strategies and decision-making. However, conventional methods are limited to mere data collection and analysis, and do not lead to actual decision-making support or the provision of concrete advice. To solve this problem and realize more sophisticated decision-making support, a new system utilizing generative AI models is necessary.

[0715] 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.

[0716] In this invention, the server includes means for collecting behavioral data of past corporate managers; means for preprocessing the collected data by standardizing it into a text format, deleting unnecessary information, and supplementing missing data; means for training a generative AI model using the preprocessed data; means for sending user inquiries to the server via a terminal; means for analyzing the received inquiries, extracting key keywords, inputting them into the generative AI model, and generating answers; means for verifying the generated answers on the server, checking grammar and content appropriateness, and making adjustments as necessary; and means for sending the final answers from the server to the terminal and displaying them to the user. This makes it possible to utilize the knowledge and experience of past corporate managers and reflect their valuable opinions in current corporate management even after their retirement.

[0717] "Data on past management behavior within a company" refers to information such as presentations, speeches, meeting notes, and decision-making records made by past managers within a company, particularly the Chief Executive Officer (CEO).

[0718] "Preprocessing to unify into text format, remove unnecessary information, and fill in missing data" refers to a series of processes that convert collected data into a unified text format, remove unnecessary tags and metadata, and fill in incomplete data.

[0719] "Training a generative AI model" refers to the process of training a generative artificial intelligence model using pre-processed data to learn the CEO's language style and decision-making patterns.

[0720] "Sending user inquiries to the server via the terminal" refers to a series of steps in which questions and inquiries entered by the user using their terminal are sent to the server in a data format such as JSON.

[0721] "Analyzing received inquiries, extracting key keywords, inputting them into a generative AI model, and generating answers" refers to the process by which a server analyzes inquiries received from users using natural language processing technology, extracts important keywords, inputs them into a generative AI model, and generates appropriate answers.

[0722] "Validating the generated responses on the server, checking grammar and content appropriateness, and making adjustments as needed" refers to the process of checking the accuracy and appropriateness of the grammar and content of responses generated by generative AI models, and making manual adjustments as necessary.

[0723] "Sending the final response from the server to the terminal and displaying it to the user" refers to a series of steps in which the server sends the verified and adjusted response to the terminal, and the terminal displays the received response to the user.

[0724] This invention is a system for collecting past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructing a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[0725] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. The hardware used includes internal servers and cloud servers. The collected data is standardized into text format using scripting languages ​​such as Python and Perl, and pre-processed by removing unnecessary information and filling in missing data.

[0726] The preprocessed data is fed into a generative artificial intelligence model (e.g., TensorFlow or PyTorch). This allows the model to learn the language styles and decision-making patterns of past CEOs. For example, presentation files from the past five years are extracted from the company's presentation database, unnecessary information is removed using regular expressions, and the data is standardized into text format.

[0727] A user sends a business-related question from their device to the server. For example, they might enter a prompt such as, "Please advise me on our market strategy for the next quarter." The server analyzes the received question using natural language processing technology and extracts key keywords. Based on these keywords, a generative AI model generates a prompt and then an answer.

[0728] The generated responses are validated on the server to check grammar and the appropriateness of the content. If there are any inappropriate parts, the server manually adjusts them. For example, if the AI ​​model responds, "The market launch timing is next year," the server verifies whether that information is accurate. The final confirmed response is sent to the terminal and displayed to the user.

[0729] The responses obtained from users via their devices can be used to inform business strategies and decision-making. For example, if a user asks, "Please give me your opinion on future market strategies," the device will display specific advice such as, "Considering the actions of competitors, six months prior to market launch would be appropriate."

[0730] This system makes it possible to leverage the knowledge and experience of past company managers and incorporate their valuable opinions into current company operations even after they have retired.

[0731] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0732] Step 1:

[0733] Data collection

[0734] The server collects past administrator activity data from internal corporate databases and publicly available documents. This primarily includes presentation files, meeting notes, and decision-making records. For example, the server periodically scans presentation files stored in the company's cloud storage for the past five years to collect new data. The input is the corporate database, and the output is the collected activity data.

[0735] Step 2:

[0736] Data preprocessing

[0737] The server collects data, standardizes it into text format, removes unnecessary information using regular expressions, and fills in missing data. As a specific example, it removes HTML tags and metadata from collected presentation data and converts it to text format. The input is collected behavioral data, and the output is pre-processed text data.

[0738] Step 3:

[0739] Training an AI model

[0740] The server uses pre-processed data to train a generative AI model. Here, frameworks such as TensorFlow and PyTorch are used to learn the administrator's language style and decision-making patterns. Specifically, the server uses a GPU to process a large dataset at high speed and train the model. The input is pre-processed text data, and the output is the trained generative AI model.

[0741] Step 4:

[0742] User inquiry processing

[0743] A user sends a business-related question to the server from their device. For example, they might enter the prompt, "Please advise on our market strategy for the next quarter." The input is the user's question text, and the output is this text converted into JSON format.

[0744] Step 5:

[0745] Question analysis

[0746] The server analyzes the received question using natural language processing techniques and extracts key keywords. As a specific example, the server extracts the keywords "market strategy" and "next quarter." The input is question data in JSON format, and the output is the extracted keywords.

[0747] Step 6:

[0748] Answer generation

[0749] The server generates prompt sentences for the AI ​​model based on the extracted keywords, and then inputs them into the model to generate a response. For example, if the keywords "market strategy, next quarter" are input into the model, it will generate the response "We recommend strengthening your competitive analysis." The input is the extracted keywords, and the output is the generated response text.

[0750] Step 7:

[0751] Verification of the answer

[0752] The server checks the grammar and appropriateness of the generated response and makes adjustments as needed. For example, if there are errors or inappropriate parts in the generated response, it will be manually corrected. The input is the generated response text, and the output is the validated final response.

[0753] Step 8:

[0754] Providing a response

[0755] The server sends the verified response to the terminal and displays it to the user. For example, the response displayed on the terminal might be, "As part of our market strategy for the next quarter, please consider obtaining a patent to differentiate our product." The input is the verified final response text, and the output is the final response displayed to the user.

[0756] (Application Example 1)

[0757] 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."

[0758] Companies need to immediately leverage the knowledge and experience of past managers, especially CEOs, to support on-the-ground decision-making. However, there is a lack of effective means to utilize the knowledge and experience of retired CEOs. Furthermore, while optimization suggestions based on real-time data are essential for factory operations, a suitable system for this purpose does not exist. To solve these problems, a system is needed that models the knowledge of past CEOs and links it with actual operational data.

[0759] 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.

[0760] In this invention, the server includes means for collecting past behavioral data of a company's managers, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for collecting factory operation data in real time, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, and means for providing the processing results to the user. This enables real-time decision support and optimization suggestions based on the knowledge and experience of past managers.

[0761] "Behavioral data" refers to data that includes information on presentations, speeches, decision-making processes, and other activities that company managers have conducted in the past.

[0762] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[0763] A "generative artificial intelligence model" is an AI model that is trained using collected and pre-processed data, and it uses natural language processing techniques to generate answers to user questions.

[0764] "User inquiries" refer to questions or consultations made by individuals within a factory or company seeking advice on specific operations or market strategies.

[0765] "Methods for collecting data in real time" refers to a system that uses sensors and cameras within the factory to instantly transmit factory operation data to a server.

[0766] "Means for processing inquiries using generative artificial intelligence models" refers to processing methods for receiving inquiries from users, analyzing them using generative artificial intelligence models, and deriving appropriate answers.

[0767] "Means of providing to the user" refers to a system for verifying the answers and suggestions generated by generative artificial intelligence models and presenting them to the user in an appropriate format.

[0768] This invention is a system that collects behavioral data of past corporate managers (especially CEOs) and constructs a virtual CEO using a generative artificial intelligence model. A specific example of this system is shown below.

[0769] Data Acquisition and Preprocessing

[0770] The server collects data on past administrator presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data undergoes preprocessing, including string organization, removal of unnecessary data, and imputation of missing data. Python and the Pandas library are used for this preprocessing.

[0771] Training an AI model

[0772] The server uses a pre-processed dataset to build a generative artificial intelligence model. Training is performed using frameworks such as TensorFlow and PyTorch. This model learns from past speeches and decision-making patterns and is trained to reproduce the CEO's thought patterns.

[0773] Data acquisition equipment

[0774] Sensors and cameras are placed throughout the factory. Operational data is collected from these devices in real time and transmitted to a server.

[0775] User inquiries and response generation

[0776] Factory workers and managers can ask questions to the virtual CEO via tablet devices or voice input devices. For example, a specific question might be, "Please give us your opinion on future market strategies." This question is sent from the tablet device to the server in JSON format.

[0777] Model-based processing and response generation

[0778] As mentioned earlier, the server analyzes this query using NLP technology and inputs it into a generative artificial intelligence model. The model generates an appropriate response based on the query. For example, it might generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[0779] Providing a response

[0780] The server validates the generated response and, if appropriate, sends it to the worker's tablet device. The tablet device displays the received response to the user, who then makes a decision based on the advice.

[0781] Examples of prompt statements

[0782] The worker provides the virtual CEO with the following prompt:

[0783] The utilization rate of our factory's production lines is decreasing. Based on past data, please provide advice on how to improve this situation.

[0784] Based on this prompt, the generative artificial intelligence model generates the following response:

[0785] According to the analysis, machine A on line 1 has been inadequately maintained for the past two weeks. Increasing the maintenance frequency is expected to improve the operating rate by 20%.

[0786] In this way, the system of this invention can leverage the knowledge and experience of past managers and, in conjunction with real-time factory data, provide optimal decision-making support.

[0787] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0788] Step 1: Data Collection

[0789] The server collects data on past administrator behavior from internal corporate databases and publicly available documents. This data includes records of presentations, speeches, and decision-making. The server stores the collected data in string format.

[0790] Input: Corporate databases, publicly available documents

[0791] Output: Behavioral data in string format

[0792] Step 2: Data Preprocessing

[0793] The server preprocesses the collected behavioral data. This preprocessing includes removing unnecessary data, imputing missing data, and standardizing the text format. This is done using Python and the Pandas library.

[0794] Input: Behavioral data in string format

[0795] Output: Preprocessed text data

[0796] Step 3: Training the AI ​​model

[0797] The server uses pre-processed data to train a generative artificial intelligence model. Using TensorFlow or PyTorch, the model learns the administrator's thought patterns and decision-making patterns.

[0798] Input: Preprocessed text data

[0799] Output: Trained generative AI model

[0800] Step 4: Real-time data collection

[0801] Sensors and cameras within the factory collect operational data in real time and transmit it to a server. This data includes machine operating status, temperature, humidity, and other information.

[0802] Input: Sensors and cameras within the factory

[0803] Output: Real-time operation data

[0804] Step 5: User Inquiry

[0805] Users can ask questions to the virtual CEO using tablet devices or voice input devices. For example, specific questions could include, "What are your thoughts on future market strategies?"

[0806] Input: User's question (e.g., "Please give us your opinion on future market strategies")

[0807] Output: Text data of the question

[0808] Step 6: Analyze the inquiry

[0809] The server analyzes the user's question and extracts key keywords using NLP techniques. These keywords are then input into a generative artificial intelligence model.

[0810] Input: Text data of the question

[0811] Output: Keyword extraction results using NLP technology

[0812] Step 7: Answer generation using AI model

[0813] Generative artificial intelligence models generate appropriate answers based on analyzed user questions. For example, they can generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[0814] Input: Keyword extraction results

[0815] Output: Text data of the response

[0816] Step 8: Verification of the answer

[0817] The server grammatically checks the generated responses and verifies their content validity. Once verification is complete, it selects the appropriate response and makes it the final answer.

[0818] Input: Text data of the response

[0819] Output: Verified text data

[0820] Step 9: Provide your response

[0821] The server sends verified answers to the user's tablet device. The device displays the received answers to the user. The user makes a decision based on the displayed advice.

[0822] Input: Verified text data

[0823] Output: Display of answers on a tablet device

[0824] In this way, it becomes possible to leverage the knowledge and experience of past managers and integrate it with real-time factory data to provide optimal decision-making support.

[0825] 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.

[0826] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[0827] System configuration and program processing

[0828] Data Acquisition and AI Model Training

[0829] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[0830] User inquiry processing and sentiment recognition

[0831] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of the market launch of a new product." The terminal has an emotion engine built in, which analyzes not only the user's input but also their emotional state at the time of input. The emotion engine detects the user's emotional state using voice analysis and facial recognition technology and sends the analysis results to the server.

[0832] Query and sentiment data processing

[0833] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate answers. The generative AI model generates answers while also considering the user's emotional state.

[0834] Generating and providing answers

[0835] The AI ​​model's generated responses are validated by the server. This validation includes checking the appropriateness and grammar of the responses. The generated responses are adjusted as needed. Finally, the server sends the finalized response to the device. The device then displays the received response to the user. For example, if the user is nervous, the response might be delivered in a softer tone, such as, "We believe that six months before your competitors' release dates is the appropriate time to launch your new product, but please let us know if you have any further questions."

[0836] Specific examples

[0837] Data Acquisition and Preprocessing

[0838] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[0839] Training an AI model

[0840] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as Tensorflow and PyTorch. During the training process, the model learns from past CEO speeches and decision-making examples to enable it to reproduce the thought patterns of CEOs.

[0841] User sentiment recognition and inquiries

[0842] The user types "Please give me your opinion on future market strategies" from their device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone, detecting their emotional state, such as whether they are tense or relaxed. The query data, including the analysis results, is sent to the server in JSON format.

[0843] Analysis and response generation on the server

[0844] Based on the data received by the server, it queries a generative AI model. The AI ​​model, taking into account the user's emotional state, generates a response such as, "Considering the actions of competitors, the optimal timing for market launch is six months ago."

[0845] Providing a response

[0846] The server verifies the generated response and, if appropriate, sends it directly to the device. The device displays the received response to the user, who then makes a decision based on the advice. For example, if the user is feeling anxious, the device might display a gentler message such as, "Set the market launch date to six months before your competitors' release schedule."

[0847] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby realizing a more human-like conversational experience.

[0848] The following describes the processing flow.

[0849] Step 1:

[0850] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also collects data from the internet and other internal systems using APIs and scraping tools.

[0851] Step 2:

[0852] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. The server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[0853] Step 3:

[0854] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[0855] Step 4:

[0856] Users input business inquiries and questions through the terminal's interface. For example, they might input, "Please advise me on the timing of the market launch for a new product."

[0857] Step 5:

[0858] The device's camera and microphone analyze the user's facial expressions and voice using an emotion engine to detect their emotional state. The device processes the analysis results in real time on the client side and sends them to the server along with the question data.

[0859] Step 6:

[0860] The device sends the user's questions and sentiment data to the server in JSON format. This transmission is done using an HTTP request.

[0861] Step 7:

[0862] The server analyzes the questions and sentiment data received from the user. Natural language processing (NLP) techniques are used to analyze the question content, extract key keywords, and classify them into appropriate themes and categories.

[0863] Step 8:

[0864] The server inputs questions and sentiment data into a generative AI model and instructs it to generate answers. The AI ​​model generates specific and contextually relevant answers while considering the user's emotional state. For example, it might generate an answer such as, "The optimal timing for market launch is six months before your competitors' release plans."

[0865] Step 9:

[0866] The server validates the generated response. This validation includes checking the appropriateness and grammar of the response. The server adjusts the content of the response as needed.

[0867] Step 10:

[0868] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[0869] Step 11:

[0870] The device displays the received response to the user. The emotion engine takes into account the user's emotional state, which it has detected in advance, and displays the response in an appropriate tone and wording. For example, if the user is nervous, the device will display in a gentle tone, "We recommend setting your market launch date six months before your competitors' release dates."

[0871] In this way, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience, and makes the conversation more human-like by taking the user's emotions into consideration.

[0872] (Example 2)

[0873] 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."

[0874] In modern business management, rapid and appropriate decision-making is essential. In particular, decision-making processes that leverage the CEO's past knowledge and experience are highly valuable. However, a system that integrates data collection, analysis, and response generation that considers user emotions does not exist. Furthermore, conventional artificial intelligence systems have struggled to generate responses that take user emotional states into account. This limits the user experience, highlighting the need for a system that can provide appropriate decision-making support.

[0875] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past behavioral data of the company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving user inquiries via a terminal, means for analyzing the received inquiries and classifying them into appropriate themes or categories, means including an emotion engine that analyzes the user's emotional state via the terminal at the time of inquiry, means for inputting the analyzed emotional state and inquiry content into the generative artificial intelligence model and processing it, means for verifying and adjusting the processing results, and means for providing the processing results to the user via the terminal. This makes it possible to provide a function as a virtual advisor that leverages the company's past knowledge and experience, and also enables the generation of detailed responses according to the user's emotional state.

[0876] A "company" is a legal entity or business that conducts economic activities as an organization.

[0877] A "manager" is a person who is responsible for effectively managing and operating operations, personnel, and resources within a company.

[0878] "Behavioral data" refers to information about work-related actions, such as managers' decision-making, statements made in meetings, and official documents.

[0879] "Preprocessing" refers to processes performed after data collection, such as deleting unnecessary information and standardizing data formats.

[0880] A "generative artificial intelligence model" is an artificial intelligence technology that learns from past data and generates responses in natural language for new data.

[0881] A "user" is an individual or organization that uses this system to receive information or support.

[0882] "Inquiry" refers to questions or inquiries entered by users into the system.

[0883] A "terminal" is a hardware device, such as a computer or smartphone, used to access and operate a system.

[0884] An "emotion engine" is a software technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state.

[0885] "Verification" refers to the process of checking whether the generated response is appropriate and grammatically correct.

[0886] "Providing" refers to the act of a server displaying or transmitting a generated response to a user.

[0887] This invention relates to a system that collects behavioral data from past corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[0888] Specific examples

[0889] Data Acquisition and Preprocessing

[0890] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data is standardized into a text format. As a preprocessing step, regular expressions are used to remove unnecessary information and standardize the data format.

[0891] Specific example: A server extracts CEO speeches, meeting notes, and decision-making records from the company's internal database over the past five years. After collection, it uses regular expressions to remove unnecessary metadata.

[0892] Training an AI model

[0893] The server uses frameworks such as TensorFlow and PyTorch to build a generative artificial intelligence model using a preprocessed dataset. By training this model with the behavioral patterns and language styles of past CEOs, a virtual CEO is created.

[0894] Specific example: The server defines a neural network using TensorFlow and trains it for 100 epochs using historical datasets.

[0895] User inquiry processing and sentiment recognition

[0896] Users input specific business inquiries and questions through the terminal's interface. The terminal is equipped with an emotion engine, which analyzes not only the user's input but also their emotional state at the time of input.

[0897] Specific example: The user types "Please give me your opinion on future market strategies" into the device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone to determine whether the user is tense or relaxed.

[0898] Query and sentiment data processing

[0899] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions and classifies them into appropriate themes and categories. It then inputs this data into a generative artificial intelligence model to generate responses that take the user's emotional state into account.

[0900] Specific example: The terminal sends the message "Please advise me on the timing of market launch" and emotion data indicating "tension" which are received by the server. The server categorizes the question under the theme of "market launch timing" and inputs it into a generative artificial intelligence model.

[0901] Generating and providing answers

[0902] The responses generated by the generative artificial intelligence model are validated by the server. This validation includes checking the appropriateness and grammar of the responses, and adjustments are made as needed. Finally, the server sends the finalized response to the terminal. The terminal then displays the received response to the user.

[0903] Specific example: The AI ​​model generates the response, "The optimal timing for market launch is six months in advance," which is validated by the server. After checking for appropriateness and grammar, the final response, "The appropriate time for market launch is six months before the competitor's planned release," is sent to the terminal. The terminal then displays this response to the user.

[0904] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby achieving a more human-like interaction.

[0905] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0906] Step 1: Data Collection

[0907] The server collects data on past CEO presentations, speeches, and decision-making from internal company databases and publicly available documents. Input is from internal company databases and online documents, from which data such as speeches, meeting notes, and decision-making records are extracted. The output is the collected, unprocessed behavioral data.

[0908] Step 2: Data preprocessing

[0909] The server converts the collected data into a unified text format and removes unnecessary information. The input is the raw behavioral data collected in step 1. The server uses regular expressions to remove unnecessary metadata and unify the data format into a consistent format. The output is the pre-processed text data.

[0910] Step 3: Training the AI ​​model

[0911] The server trains a generative artificial intelligence model using a preprocessed dataset. The input is the preprocessed text data obtained in step 2. Specifically, the model is built using TensorFlow or PyTorch and trained using the dataset. The output is the trained generative artificial intelligence model.

[0912] Step 4: Receiving user inquiry

[0913] The user enters specific business inquiries or questions through the terminal's interface. The input is a text-based question (prompt) entered by the user. The terminal receives this input and proceeds to the next step. The output is the user's inquiry text.

[0914] Step 5: Analysis of emotional state

[0915] The emotion engine analyzes the user's emotional state through the device's camera and microphone. The input is the user's facial expressions and tone of voice. The device uses facial recognition and voice analysis technologies to analyze the user's emotional state and extract emotional characteristics such as tension and relaxation. The output is the analyzed user's emotional state data.

[0916] Step 6: Data transmission

[0917] The terminal sends the user's inquiry and sentiment data to the server. The input is the user's inquiry text obtained in step 4 and the sentiment state data obtained in step 5. The terminal converts this data into JSON format and sends it to the server. The output is the JSON data sent to the server.

[0918] Step 7: Question Analysis and Categorization

[0919] The server analyzes the received data and classifies the questions into appropriate themes and categories. The input is the JSON data received from the terminal in step 6. The server uses natural language processing technology to analyze the questions and classify them into themes such as "timing of market entry" and "competitive analysis." The output is the classified question data.

[0920] Step 8: Processing Questions and Sentimental Data

[0921] The server inputs question and emotion state data into a generative artificial intelligence model and instructs it to generate an answer. The input consists of the question data classified in step 7 and the emotion state data received in step 6. The generative artificial intelligence model generates an answer based on this data. The output is the answer generated by the AI ​​model.

[0922] Step 9: Verify and adjust your response

[0923] The server validates the generated response and adjusts it as needed. The input is the response generated in step 8. The server checks the appropriateness and grammar of the response and adjusts the content if necessary. The output is the validated and adjusted response.

[0924] Step 10: Provide your response

[0925] The server provides the answer to the user via the terminal. The input is the answer that was verified and adjusted in step 9. The server sends the finalized answer to the terminal, and the terminal displays the received answer to the user. The output is the answer displayed to the user.

[0926] (Application Example 2)

[0927] 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."

[0928] Traditionally, artificial intelligence models generated using behavioral data from corporate administrators could produce appropriate responses to inquiries, but they could not provide responses that took into account the user's emotional state. As a result, users, especially those experiencing tension or stress, were unable to receive appropriate advice, leading to a decline in the quality of decision-making.

[0929] 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.

[0930] In this invention, the server includes means for collecting past behavioral data of a company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, means for providing the processing results to the user, means for detecting the user's emotional state using an emotion engine, and means for reflecting the user's emotional state in the response generation of the generative artificial intelligence model. This makes it possible to provide responses that take the user's emotional state into consideration.

[0931] "Data on past management behavior within a company" refers to information including records of presentations, speeches, and decision-making activities previously conducted by managers within the company.

[0932] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[0933] A "generative artificial intelligence model" is a type of artificial intelligence that is trained using collected and pre-processed data to reproduce past behavioral patterns and decision-making of administrators.

[0934] An "emotion engine" is a device or software that includes technology to analyze facial expressions and voice in order to detect the user's emotional state.

[0935] An "inquiry" is the act of a user entering specific questions or requests for advice regarding a business.

[0936] "Processing result" refers to the response content generated based on the generative artificial intelligence model and the detected emotional state.

[0937] "Reflection" refers to the process by which a generative artificial intelligence model adjusts its response in consideration of the detected emotional state of the user.

[0938] This invention is a system that collects past behavioral data of corporate managers and constructs a virtual manager using a generative artificial intelligence model. Furthermore, by combining it with an emotion engine that recognizes the user's emotional state, it can provide more human-like responses. A specific embodiment of this system is described below.

[0939] Data Acquisition and Preprocessing

[0940] The server collects data from the company's database regarding past administrator presentations, speeches, and decision-making. This data undergoes preprocessing, including standardization to text format using regular expressions, removal of unnecessary information, and imputation of missing data. The preprocessed data is then used as training data for generative artificial intelligence models.

[0941] Training of generative AI models

[0942] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow and PyTorch. This training process analyzes past administrator behavior data to learn patterns of speech and decision-making styles.

[0943] Inquiry and Sentiment Recognition

[0944] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. These devices are equipped with an emotion engine that uses the camera and microphone to analyze the user's facial expressions and tone of voice, detecting their emotional state. For example, when a user inputs "Please tell me how to resolve the bottleneck in the production line," the emotion engine analyzes whether the user is tense or relaxed.

[0945] Query and sentiment data processing

[0946] The user's inquiry and emotional data are sent to the server in JSON format. The server analyzes the received data and inputs it into a generative artificial intelligence model. This model generates responses while taking the user's emotional state into consideration. For example, if the user is nervous, the response will be provided in a gentle tone.

[0947] Generating and providing answers

[0948] The AI ​​model's generated responses are validated by the server, and if appropriate, are sent directly to the user's device. The device then displays the received responses to the user. This entire process allows the user to receive appropriate advice from a virtual administrator. For example, specific responses such as, "To eliminate bottlenecks in the production line, it would be effective to review the worker shift schedule and utilize production management software," are provided.

[0949] Specific examples and prompt statements

[0950] As a concrete example, production reports from the past 10 years can be extracted from a factory's project logs. The user can then ask questions such as, "Please tell me specific ways to improve production efficiency," and the system will recognize their facial expressions using a camera and send the results to a server, providing appropriate advice. This allows the user to take concrete measures to improve productivity.

[0951] Example of a prompt

[0952] "Please tell me about specific measures to improve production efficiency."

[0953] "I'd like some advice on the timing of market launch."

[0954] "How can we optimize claims processing for a new production line?"

[0955] This invention makes it possible to provide appropriate decision-making support that takes into account the user's emotional state while utilizing the knowledge and experience of past administrators.

[0956] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0957] Step 1:

[0958] The server collects data from the company's database regarding past administrator presentations, speeches, and decisions. This involves a data collection process that executes database queries to extract the necessary information and convert it into an appropriate format. The input is the company's database, and the output is the collected raw data.

[0959] Step 2:

[0960] The server preprocesses the collected data. In this step, unnecessary information is removed using regular expressions, missing data is imputed, and the data is standardized to text format. This allows generative artificial intelligence models to efficiently learn from the data. The input is the collected raw data, and the output is the preprocessed clean data.

[0961] Step 3:

[0962] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow or PyTorch. During this process, the data is used as a training set for the model, learning the administrator's behavioral patterns and decision-making processes. The input is pre-processed clean data, and the output is the trained AI model.

[0963] Step 4:

[0964] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. The user's input is presented as natural language prompts. The input is a text query from the user, and the output is data sent from the device to the server.

[0965] Step 5:

[0966] The emotion engine built into the device uses the camera and microphone to analyze the user's facial expressions and tone of voice to detect their emotional state. This emotion data is sent to the server in JSON format. The input is the user's voice and video data, and the output is the analyzed emotion data.

[0967] Step 6:

[0968] The server analyzes the received question content and sentiment data, and inputs it into a generative artificial intelligence model. The model considers the user's emotional state and generates an appropriate response. The input is the received question content and sentiment data, and the output is the response generated by the model.

[0969] Step 7:

[0970] The server validates the generated response and makes corrections if necessary. This validation includes checking the appropriateness and grammar of the response. The input is the generated response, and the output is the validated or corrected response.

[0971] Step 8:

[0972] The server sends the final answer to the terminal. The terminal displays the received answer to the user. This sequence of actions allows the user to receive appropriate advice from the virtual administrator. The input is the verified or corrected answer, and the output is the final answer presented to the user.

[0973] 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.

[0974] 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.

[0975] 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.

[0976] [Fourth Embodiment]

[0977] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0978] 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.

[0979] 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).

[0980] 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.

[0981] 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.

[0982] 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).

[0983] 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.

[0984] 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.

[0985] 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.

[0986] 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.

[0987] 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.

[0988] 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.

[0989] 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".

[0990] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[0991] System configuration and program processing

[0992] Data Acquisition and AI Model Training

[0993] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[0994] User inquiry processing

[0995] Users input business-related questions and inquiries from their terminals to the server. For example, specific questions might include, "Please advise on the timing of the market launch of a new product." The questions sent from the terminals are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate an answer.

[0996] Generating and providing answers

[0997] The answers generated by the AI ​​model are validated by the server. This validation includes checking the appropriateness and grammar of the answers. The generated answers are adjusted as needed. Finally, the server sends the finalized answer to the device. The device then displays the received answer to the user. For example, if the user asks, "Give me some advice on the market strategy for a new product," the device might display specific advice such as, "We believe the optimal time to launch a new product is six months before competitors plan to release theirs."

[0998] Specific examples

[0999] Data Acquisition and Preprocessing

[1000] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[1001] Training an AI model

[1002] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as Tensorflow and PyTorch. During the training process, the model learns from past speeches and decision-making examples to enable it to reproduce the CEO's thought patterns.

[1003] User inquiries and response generation

[1004] The user inputs "Please give me your opinion on future market strategies" from their device. The device sends this input to the server in JSON format. The server receives this question, extracts key keywords using NLP technology, and queries a generative AI model. The AI ​​model generates a response such as "Considering the actions of competitors, the optimal timing for market entry is six months ago."

[1005] Providing a response

[1006] The server verifies the generated response and, if appropriate, sends it directly to the terminal. The terminal displays the received response to the user, who then makes a decision based on the advice.

[1007] Thus, the system of the present invention provides a means to utilize the knowledge and experience of a company's past CEOs and to reflect their valuable opinions even after their retirement.

[1008] The following describes the processing flow.

[1009] Step 1:

[1010] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also uses APIs and scraping tools to collect data from the internet and other internal systems.

[1011] Step 2:

[1012] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. For example, the server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[1013] Step 3:

[1014] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[1015] Step 4:

[1016] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of launching a new product into the market."

[1017] Step 5:

[1018] The device sends user questions to the server in real time. During transmission, the question data is structured using JSON format, and the data is sent to the server using an HTTP request.

[1019] Step 6:

[1020] The server analyzes the questions received from users and classifies them into appropriate themes and categories. During this process, natural language processing (NLP) techniques are used to extract the main keywords of the questions and perform semantic analysis.

[1021] Step 7:

[1022] The server inputs a question into a generative AI model based on the analysis results, and generates an answer. The AI ​​model uses its past learning to generate a specific answer. For example, it might generate an answer such as, "The optimal time to launch a new product into the market is six months before competitors plan to release theirs."

[1023] Step 8:

[1024] The server validates the generated response. The validation process checks the appropriateness and grammar of the response and adjusts the content of the response as needed.

[1025] Step 9:

[1026] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[1027] Step 10:

[1028] The device displays the received responses to the user. The user reviews specific advice and opinions through the device's interface and makes decisions based on them. For example, they might adopt a strategy such as "setting the market launch date six months before the competitor's planned release."

[1029] Thus, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience through a series of processing steps.

[1030] (Example 1)

[1031] 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".

[1032] There is a need to effectively utilize the knowledge and experience of corporate managers, especially CEOs, even after their retirement, and to reflect them in current business strategies and decision-making. However, conventional methods are limited to mere data collection and analysis, and do not lead to actual decision-making support or the provision of concrete advice. To solve this problem and realize more sophisticated decision-making support, a new system utilizing generative AI models is necessary.

[1033] 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.

[1034] In this invention, the server includes means for collecting behavioral data of past corporate managers; means for preprocessing the collected data by standardizing it into a text format, deleting unnecessary information, and supplementing missing data; means for training a generative AI model using the preprocessed data; means for sending user inquiries to the server via a terminal; means for analyzing the received inquiries, extracting key keywords, inputting them into the generative AI model, and generating answers; means for verifying the generated answers on the server, checking grammar and content appropriateness, and making adjustments as necessary; and means for sending the final answers from the server to the terminal and displaying them to the user. This makes it possible to utilize the knowledge and experience of past corporate managers and reflect their valuable opinions in current corporate management even after their retirement.

[1035] "Data on past management behavior within a company" refers to information such as presentations, speeches, meeting notes, and decision-making records made by past managers within a company, particularly the Chief Executive Officer (CEO).

[1036] "Preprocessing to unify into text format, remove unnecessary information, and fill in missing data" refers to a series of processes that convert collected data into a unified text format, remove unnecessary tags and metadata, and fill in incomplete data.

[1037] "Training a generative AI model" refers to the process of training a generative artificial intelligence model using pre-processed data to learn the CEO's language style and decision-making patterns.

[1038] "Sending user inquiries to the server via the terminal" refers to a series of steps in which questions and inquiries entered by the user using their terminal are sent to the server in a data format such as JSON.

[1039] "Analyzing received inquiries, extracting key keywords, inputting them into a generative AI model, and generating answers" refers to the process by which a server analyzes inquiries received from users using natural language processing technology, extracts important keywords, inputs them into a generative AI model, and generates appropriate answers.

[1040] "Validating the generated responses on the server, checking grammar and content appropriateness, and making adjustments as needed" refers to the process of checking the accuracy and appropriateness of the grammar and content of responses generated by generative AI models, and making manual adjustments as necessary.

[1041] "Sending the final response from the server to the terminal and displaying it to the user" refers to a series of steps in which the server sends the verified and adjusted response to the terminal, and the terminal displays the received response to the user.

[1042] This invention is a system for collecting past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructing a virtual CEO using a generative artificial intelligence model. In this system, users can ask questions to the virtual CEO via a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs.

[1043] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. The hardware used includes internal servers and cloud servers. The collected data is standardized into text format using scripting languages ​​such as Python and Perl, and pre-processed by removing unnecessary information and filling in missing data.

[1044] The preprocessed data is fed into a generative artificial intelligence model (e.g., TensorFlow or PyTorch). This allows the model to learn the language styles and decision-making patterns of past CEOs. For example, presentation files from the past five years are extracted from the company's presentation database, unnecessary information is removed using regular expressions, and the data is standardized into text format.

[1045] A user sends a business-related question from their device to the server. For example, they might enter a prompt such as, "Please advise me on our market strategy for the next quarter." The server analyzes the received question using natural language processing technology and extracts key keywords. Based on these keywords, a generative AI model generates a prompt and then an answer.

[1046] The generated responses are validated on the server to check grammar and the appropriateness of the content. If there are any inappropriate parts, the server manually adjusts them. For example, if the AI ​​model responds, "The market launch timing is next year," the server verifies whether that information is accurate. The final confirmed response is sent to the terminal and displayed to the user.

[1047] The responses obtained from users via their devices can be used to inform business strategies and decision-making. For example, if a user asks, "Please give me your opinion on future market strategies," the device will display specific advice such as, "Considering the actions of competitors, six months prior to market launch would be appropriate."

[1048] This system makes it possible to leverage the knowledge and experience of past company managers and incorporate their valuable opinions into current company operations even after they have retired.

[1049] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1050] Step 1:

[1051] Data collection

[1052] The server collects past administrator activity data from internal corporate databases and publicly available documents. This primarily includes presentation files, meeting notes, and decision-making records. For example, the server periodically scans presentation files stored in the company's cloud storage for the past five years to collect new data. The input is the corporate database, and the output is the collected activity data.

[1053] Step 2:

[1054] Data preprocessing

[1055] The server collects data, standardizes it into text format, removes unnecessary information using regular expressions, and fills in missing data. As a specific example, it removes HTML tags and metadata from collected presentation data and converts it to text format. The input is collected behavioral data, and the output is pre-processed text data.

[1056] Step 3:

[1057] Training an AI model

[1058] The server uses pre-processed data to train a generative AI model. Here, frameworks such as TensorFlow and PyTorch are used to learn the administrator's language style and decision-making patterns. Specifically, the server uses a GPU to process a large dataset at high speed and train the model. The input is pre-processed text data, and the output is the trained generative AI model.

[1059] Step 4:

[1060] User inquiry processing

[1061] A user sends a business-related question to the server from their device. For example, they might enter the prompt, "Please advise on our market strategy for the next quarter." The input is the user's question text, and the output is this text converted into JSON format.

[1062] Step 5:

[1063] Question analysis

[1064] The server analyzes the received question using natural language processing techniques and extracts key keywords. As a specific example, the server extracts the keywords "market strategy" and "next quarter." The input is question data in JSON format, and the output is the extracted keywords.

[1065] Step 6:

[1066] Answer generation

[1067] The server generates prompt sentences for the AI ​​model based on the extracted keywords, and then inputs them into the model to generate a response. For example, if the keywords "market strategy, next quarter" are input into the model, it will generate the response "We recommend strengthening your competitive analysis." The input is the extracted keywords, and the output is the generated response text.

[1068] Step 7:

[1069] Verification of the answer

[1070] The server checks the grammar and appropriateness of the generated response and makes adjustments as needed. For example, if there are errors or inappropriate parts in the generated response, it will be manually corrected. The input is the generated response text, and the output is the validated final response.

[1071] Step 8:

[1072] Providing a response

[1073] The server sends the verified response to the terminal and displays it to the user. For example, the response displayed on the terminal might be, "As part of our market strategy for the next quarter, please consider obtaining a patent to differentiate our product." The input is the verified final response text, and the output is the final response displayed to the user.

[1074] (Application Example 1)

[1075] 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".

[1076] Companies need to immediately leverage the knowledge and experience of past managers, especially CEOs, to support on-the-ground decision-making. However, there is a lack of effective means to utilize the knowledge and experience of retired CEOs. Furthermore, while optimization suggestions based on real-time data are essential for factory operations, a suitable system for this purpose does not exist. To solve these problems, a system is needed that models the knowledge of past CEOs and links it with actual operational data.

[1077] 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.

[1078] In this invention, the server includes means for collecting past behavioral data of a company's managers, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for collecting factory operation data in real time, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, and means for providing the processing results to the user. This enables real-time decision support and optimization suggestions based on the knowledge and experience of past managers.

[1079] "Behavioral data" refers to data that includes information on presentations, speeches, decision-making processes, and other activities that company managers have conducted in the past.

[1080] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[1081] A "generative artificial intelligence model" is an AI model that is trained using collected and pre-processed data, and it uses natural language processing techniques to generate answers to user questions.

[1082] "User inquiries" refer to questions or consultations made by individuals within a factory or company seeking advice on specific operations or market strategies.

[1083] "Methods for collecting data in real time" refers to a system that uses sensors and cameras within the factory to instantly transmit factory operation data to a server.

[1084] "Means for processing inquiries using generative artificial intelligence models" refers to processing methods for receiving inquiries from users, analyzing them using generative artificial intelligence models, and deriving appropriate answers.

[1085] "Means of providing to the user" refers to a system for verifying the answers and suggestions generated by generative artificial intelligence models and presenting them to the user in an appropriate format.

[1086] This invention is a system that collects behavioral data of past corporate managers (especially CEOs) and constructs a virtual CEO using a generative artificial intelligence model. A specific example of this system is shown below.

[1087] Data Acquisition and Preprocessing

[1088] The server collects data on past administrator presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data undergoes preprocessing, including string organization, removal of unnecessary data, and imputation of missing data. Python and the Pandas library are used for this preprocessing.

[1089] Training an AI model

[1090] The server uses a pre-processed dataset to build a generative artificial intelligence model. Training is performed using frameworks such as TensorFlow and PyTorch. This model learns from past speeches and decision-making patterns and is trained to reproduce the CEO's thought patterns.

[1091] Data acquisition equipment

[1092] Sensors and cameras are placed throughout the factory. Operational data is collected from these devices in real time and transmitted to a server.

[1093] User inquiries and response generation

[1094] Factory workers and managers can ask questions to the virtual CEO via tablet devices or voice input devices. For example, a specific question might be, "Please give us your opinion on future market strategies." This question is sent from the tablet device to the server in JSON format.

[1095] Model-based processing and response generation

[1096] As mentioned earlier, the server analyzes this query using NLP technology and inputs it into a generative artificial intelligence model. The model generates an appropriate response based on the query. For example, it might generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[1097] Providing a response

[1098] The server validates the generated response and, if appropriate, sends it to the worker's tablet device. The tablet device displays the received response to the user, who then makes a decision based on the advice.

[1099] Examples of prompt statements

[1100] The worker provides the virtual CEO with the following prompt:

[1101] The utilization rate of our factory's production lines is decreasing. Based on past data, please provide advice on how to improve this situation.

[1102] Based on this prompt, the generative artificial intelligence model generates the following response:

[1103] According to the analysis, machine A on line 1 has been inadequately maintained for the past two weeks. Increasing the maintenance frequency is expected to improve the operating rate by 20%.

[1104] In this way, the system of this invention can leverage the knowledge and experience of past managers and, in conjunction with real-time factory data, provide optimal decision-making support.

[1105] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1106] Step 1: Data Collection

[1107] The server collects data on past administrator behavior from internal corporate databases and publicly available documents. This data includes records of presentations, speeches, and decision-making. The server stores the collected data in string format.

[1108] Input: Corporate databases, publicly available documents

[1109] Output: Behavioral data in string format

[1110] Step 2: Data Preprocessing

[1111] The server preprocesses the collected behavioral data. This preprocessing includes removing unnecessary data, imputing missing data, and standardizing the text format. This is done using Python and the Pandas library.

[1112] Input: Behavioral data in string format

[1113] Output: Preprocessed text data

[1114] Step 3: Training the AI ​​model

[1115] The server uses pre-processed data to train a generative artificial intelligence model. Using TensorFlow or PyTorch, the model learns the administrator's thought patterns and decision-making patterns.

[1116] Input: Preprocessed text data

[1117] Output: Trained generative AI model

[1118] Step 4: Real-time data collection

[1119] Sensors and cameras within the factory collect operational data in real time and transmit it to a server. This data includes machine operating status, temperature, humidity, and other information.

[1120] Input: Sensors and cameras within the factory

[1121] Output: Real-time operation data

[1122] Step 5: User Inquiry

[1123] Users can ask questions to the virtual CEO using tablet devices or voice input devices. For example, specific questions could include, "What are your thoughts on future market strategies?"

[1124] Input: User's question (e.g., "Please give us your opinion on future market strategies")

[1125] Output: Text data of the question

[1126] Step 6: Analyze the inquiry

[1127] The server analyzes the user's question and extracts key keywords using NLP techniques. These keywords are then input into a generative artificial intelligence model.

[1128] Input: Text data of the question

[1129] Output: Keyword extraction results using NLP technology

[1130] Step 7: Answer generation using AI model

[1131] Generative artificial intelligence models generate appropriate answers based on analyzed user questions. For example, they can generate specific advice such as, "Considering the actions of competitors, the optimal timing for market launch is six months in advance."

[1132] Input: Keyword extraction results

[1133] Output: Text data of the response

[1134] Step 8: Verification of the answer

[1135] The server grammatically checks the generated responses and verifies their content validity. Once verification is complete, it selects the appropriate response and makes it the final answer.

[1136] Input: Text data of the response

[1137] Output: Verified text data

[1138] Step 9: Provide your response

[1139] The server sends verified answers to the user's tablet device. The device displays the received answers to the user. The user makes a decision based on the displayed advice.

[1140] Input: Verified text data

[1141] Output: Display of answers on a tablet device

[1142] In this way, it becomes possible to leverage the knowledge and experience of past managers and integrate it with real-time factory data to provide optimal decision-making support.

[1143] 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.

[1144] This invention relates to a system that collects past behavioral data of corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[1145] System configuration and program processing

[1146] Data Acquisition and AI Model Training

[1147] A server collects data on past CEO presentations, speeches, and decision-making from corporate databases and publicly available documents. This data is standardized into text format and pre-processed. Pre-processing includes removing unnecessary information and filling in missing data. The pre-processed data is fed into a generative artificial intelligence model to train a virtual CEO. This model learns the CEO's language style and decision-making patterns.

[1148] User inquiry processing and sentiment recognition

[1149] Users input specific business inquiries or questions through the terminal's interface. For example, a user might input, "Please advise me on the timing of the market launch of a new product." The terminal has an emotion engine built in, which analyzes not only the user's input but also their emotional state at the time of input. The emotion engine detects the user's emotional state using voice analysis and facial recognition technology and sends the analysis results to the server.

[1150] Query and sentiment data processing

[1151] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions, classifies them into appropriate themes and categories, and then inputs them into a generative artificial intelligence model, instructing it to generate answers. The generative AI model generates answers while also considering the user's emotional state.

[1152] Generating and providing answers

[1153] The AI ​​model's generated responses are validated by the server. This validation includes checking the appropriateness and grammar of the responses. The generated responses are adjusted as needed. Finally, the server sends the finalized response to the device. The device then displays the received response to the user. For example, if the user is nervous, the response might be delivered in a softer tone, such as, "We believe that six months before your competitors' release dates is the appropriate time to launch your new product, but please let us know if you have any further questions."

[1154] Specific examples

[1155] Data Acquisition and Preprocessing

[1156] The server extracts CEO speeches, meeting notes, and decision-making records from the past five years from the company's internal presentation database. After collection, unnecessary information is removed using regular expressions, and the data is standardized to text format.

[1157] Training an AI model

[1158] The server uses a pre-processed dataset to build a generative artificial intelligence model using frameworks such as Tensorflow and PyTorch. During the training process, the model learns from past CEO speeches and decision-making examples to enable it to reproduce the thought patterns of CEOs.

[1159] User sentiment recognition and inquiries

[1160] The user types "Please give me your opinion on future market strategies" from their device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone, detecting their emotional state, such as whether they are tense or relaxed. The query data, including the analysis results, is sent to the server in JSON format.

[1161] Analysis and response generation on the server

[1162] Based on the data received by the server, it queries a generative AI model. The AI ​​model, taking into account the user's emotional state, generates a response such as, "Considering the actions of competitors, the optimal timing for market launch is six months ago."

[1163] Providing a response

[1164] The server verifies the generated response and, if appropriate, sends it directly to the device. The device displays the received response to the user, who then makes a decision based on the advice. For example, if the user is feeling anxious, the device might display a gentler message such as, "Set the market launch date to six months before your competitors' release schedule."

[1165] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby realizing a more human-like conversational experience.

[1166] The following describes the processing flow.

[1167] Step 1:

[1168] The server collects data on past CEO presentations, speeches, and decision-making from the company's databases and publicly available documents. Specifically, the server executes queries such as "SELECT FROM CEO_PRESENTATIONS" to extract the necessary information from the database. It also collects data from the internet and other internal systems using APIs and scraping tools.

[1169] Step 2:

[1170] The server preprocesses the collected data. Preprocessing includes removing unnecessary information, standardizing the format, and cleaning up the text. The server uses Python's regular expression module to remove noise and the Pandas library to impute missing data. This step generates a preprocessed dataset.

[1171] Step 3:

[1172] The server feeds a pre-processed dataset into a generative artificial intelligence model and trains the model. Specifically, the server builds the model using machine learning frameworks such as Tensorflow and PyTorch. During training, it learns the language styles and decision-making patterns of past CEOs, gradually improving the generative AI model.

[1173] Step 4:

[1174] Users input business inquiries and questions through the terminal's interface. For example, they might input, "Please advise me on the timing of the market launch for a new product."

[1175] Step 5:

[1176] The device's camera and microphone analyze the user's facial expressions and voice using an emotion engine to detect their emotional state. The device processes the analysis results in real time on the client side and sends them to the server along with the question data.

[1177] Step 6:

[1178] The device sends the user's questions and sentiment data to the server in JSON format. This transmission is done using an HTTP request.

[1179] Step 7:

[1180] The server analyzes the questions and sentiment data received from the user. Natural language processing (NLP) techniques are used to analyze the question content, extract key keywords, and classify them into appropriate themes and categories.

[1181] Step 8:

[1182] The server inputs questions and sentiment data into a generative AI model and instructs it to generate answers. The AI ​​model generates specific and contextually relevant answers while considering the user's emotional state. For example, it might generate an answer such as, "The optimal timing for market launch is six months before your competitors' release plans."

[1183] Step 9:

[1184] The server validates the generated response. This validation includes checking the appropriateness and grammar of the response. The server adjusts the content of the response as needed.

[1185] Step 10:

[1186] The server sends the confirmed response to the device. The response is again structured in JSON format and sent to the device via an HTTP response.

[1187] Step 11:

[1188] The device displays the received response to the user. The emotion engine takes into account the user's emotional state, which it has detected in advance, and displays the response in an appropriate tone and wording. For example, if the user is nervous, the device will display in a gentle tone, "We recommend setting your market launch date six months before your competitors' release dates."

[1189] In this way, the system of the present invention provides a virtual advisor that leverages the CEO's past knowledge and experience, and makes the conversation more human-like by taking the user's emotions into consideration.

[1190] (Example 2)

[1191] 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".

[1192] In modern business management, rapid and appropriate decision-making is essential. In particular, decision-making processes that leverage the CEO's past knowledge and experience are highly valuable. However, a system that integrates data collection, analysis, and response generation that considers user emotions does not exist. Furthermore, conventional artificial intelligence systems have struggled to generate responses that take user emotional states into account. This limits the user experience, highlighting the need for a system that can provide appropriate decision-making support.

[1193] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting past behavioral data of the company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving user inquiries via a terminal, means for analyzing the received inquiries and classifying them into appropriate themes or categories, means including an emotion engine that analyzes the user's emotional state via the terminal at the time of inquiry, means for inputting the analyzed emotional state and inquiry content into the generative artificial intelligence model and processing it, means for verifying and adjusting the processing results, and means for providing the processing results to the user via the terminal. This makes it possible to provide a function as a virtual advisor that leverages the company's past knowledge and experience, and also enables the generation of detailed responses according to the user's emotional state.

[1194] A "company" is a legal entity or business that conducts economic activities as an organization.

[1195] A "manager" is a person who is responsible for effectively managing and operating operations, personnel, and resources within a company.

[1196] "Behavioral data" refers to information about work-related actions, such as managers' decision-making, statements made in meetings, and official documents.

[1197] "Preprocessing" refers to processes performed after data collection, such as deleting unnecessary information and standardizing data formats.

[1198] A "generative artificial intelligence model" is an artificial intelligence technology that learns from past data and generates responses in natural language for new data.

[1199] A "user" is an individual or organization that uses this system to receive information or support.

[1200] "Inquiry" refers to questions or inquiries entered by users into the system.

[1201] A "terminal" is a hardware device, such as a computer or smartphone, used to access and operate a system.

[1202] An "emotion engine" is a software technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state.

[1203] "Verification" refers to the process of checking whether the generated response is appropriate and grammatically correct.

[1204] "Providing" refers to the act of a server displaying or transmitting a generated response to a user.

[1205] This invention relates to a system that collects behavioral data from past corporate managers, particularly chief executive officers (CEOs), and constructs a virtual CEO using a generative artificial intelligence model, further incorporating an emotion engine that recognizes the user's emotions. In this system, users can ask questions to the virtual CEO through a server, receiving advice and decision-making support based on the knowledge and experience of past CEOs, while the emotion engine recognizes the user's emotional state and provides appropriate responses that reflect it.

[1206] Specific examples

[1207] Data Acquisition and Preprocessing

[1208] The server collects data on past CEO presentations, speeches, and decision-making from the company's internal databases and publicly available documents. This collected data is standardized into a text format. As a preprocessing step, regular expressions are used to remove unnecessary information and standardize the data format.

[1209] Specific example: A server extracts CEO speeches, meeting notes, and decision-making records from the company's internal database over the past five years. After collection, it uses regular expressions to remove unnecessary metadata.

[1210] Training an AI model

[1211] The server uses frameworks such as TensorFlow and PyTorch to build a generative artificial intelligence model using a preprocessed dataset. By training this model with the behavioral patterns and language styles of past CEOs, a virtual CEO is created.

[1212] Specific example: The server defines a neural network using TensorFlow and trains it for 100 epochs using historical datasets.

[1213] User inquiry processing and sentiment recognition

[1214] Users input specific business inquiries and questions through the terminal's interface. The terminal is equipped with an emotion engine, which analyzes not only the user's input but also their emotional state at the time of input.

[1215] Specific example: The user types "Please give me your opinion on future market strategies" into the device. The emotion engine analyzes the user's facial expressions and tone of voice through the device's camera and microphone to determine whether the user is tense or relaxed.

[1216] Query and sentiment data processing

[1217] Questions and sentiment data sent from the terminal are received by the server. The server analyzes the received questions and classifies them into appropriate themes and categories. It then inputs this data into a generative artificial intelligence model to generate responses that take the user's emotional state into account.

[1218] Specific example: The terminal sends the message "Please advise me on the timing of market launch" and emotion data indicating "tension" which are received by the server. The server categorizes the question under the theme of "market launch timing" and inputs it into a generative artificial intelligence model.

[1219] Generating and providing answers

[1220] The responses generated by the generative artificial intelligence model are validated by the server. This validation includes checking the appropriateness and grammar of the responses, and adjustments are made as needed. Finally, the server sends the finalized response to the terminal. The terminal then displays the received response to the user.

[1221] Specific example: The AI ​​model generates the response, "The optimal timing for market launch is six months in advance," which is validated by the server. After checking for appropriateness and grammar, the final response, "The appropriate time for market launch is six months before the competitor's planned release," is sent to the terminal. The terminal then displays this response to the user.

[1222] Thus, the present invention, which combines an emotion engine, provides a virtual advisor that leverages the CEO's past knowledge and experience, and further generates appropriate responses that take into account the user's emotional state, thereby achieving a more human-like interaction.

[1223] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1224] Step 1: Data Collection

[1225] The server collects data on past CEO presentations, speeches, and decision-making from internal company databases and publicly available documents. Input is from internal company databases and online documents, from which data such as speeches, meeting notes, and decision-making records are extracted. The output is the collected, unprocessed behavioral data.

[1226] Step 2: Data preprocessing

[1227] The server converts the collected data into a unified text format and removes unnecessary information. The input is the raw behavioral data collected in step 1. The server uses regular expressions to remove unnecessary metadata and unify the data format into a consistent format. The output is the pre-processed text data.

[1228] Step 3: Training the AI ​​model

[1229] The server trains a generative artificial intelligence model using a preprocessed dataset. The input is the preprocessed text data obtained in step 2. Specifically, the model is built using TensorFlow or PyTorch and trained using the dataset. The output is the trained generative artificial intelligence model.

[1230] Step 4: Receiving user inquiry

[1231] The user enters specific business inquiries or questions through the terminal's interface. The input is a text-based question (prompt) entered by the user. The terminal receives this input and proceeds to the next step. The output is the user's inquiry text.

[1232] Step 5: Analysis of emotional state

[1233] The emotion engine analyzes the user's emotional state through the device's camera and microphone. The input is the user's facial expressions and tone of voice. The device uses facial recognition and voice analysis technologies to analyze the user's emotional state and extract emotional characteristics such as tension and relaxation. The output is the analyzed user's emotional state data.

[1234] Step 6: Data transmission

[1235] The terminal sends the user's inquiry and sentiment data to the server. The input is the user's inquiry text obtained in step 4 and the sentiment state data obtained in step 5. The terminal converts this data into JSON format and sends it to the server. The output is the JSON data sent to the server.

[1236] Step 7: Question Analysis and Categorization

[1237] The server analyzes the received data and classifies the questions into appropriate themes and categories. The input is the JSON data received from the terminal in step 6. The server uses natural language processing technology to analyze the questions and classify them into themes such as "timing of market entry" and "competitive analysis." The output is the classified question data.

[1238] Step 8: Processing Questions and Sentimental Data

[1239] The server inputs question and emotion state data into a generative artificial intelligence model and instructs it to generate an answer. The input consists of the question data classified in step 7 and the emotion state data received in step 6. The generative artificial intelligence model generates an answer based on this data. The output is the answer generated by the AI ​​model.

[1240] Step 9: Verify and adjust your response

[1241] The server validates the generated response and adjusts it as needed. The input is the response generated in step 8. The server checks the appropriateness and grammar of the response and adjusts the content if necessary. The output is the validated and adjusted response.

[1242] Step 10: Provide your response

[1243] The server provides the answer to the user via the terminal. The input is the answer that was verified and adjusted in step 9. The server sends the finalized answer to the terminal, and the terminal displays the received answer to the user. The output is the answer displayed to the user.

[1244] (Application Example 2)

[1245] 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 robot 414 as the "terminal".

[1246] Traditionally, artificial intelligence models generated using behavioral data from corporate administrators could produce appropriate responses to inquiries, but they could not provide responses that took into account the user's emotional state. As a result, users, especially those experiencing tension or stress, were unable to receive appropriate advice, leading to a decline in the quality of decision-making.

[1247] 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.

[1248] In this invention, the server includes means for collecting past behavioral data of a company's administrators, means for preprocessing the collected data, means for training a generative artificial intelligence model using the preprocessed data, means for receiving inquiries from users, means for processing the received inquiries with the generative artificial intelligence model, means for providing the processing results to the user, means for detecting the user's emotional state using an emotion engine, and means for reflecting the user's emotional state in the response generation of the generative artificial intelligence model. This makes it possible to provide responses that take the user's emotional state into consideration.

[1249] "Data on past management behavior within a company" refers to information including records of presentations, speeches, and decision-making activities previously conducted by managers within the company.

[1250] "Preprocessing" refers to the process of removing unnecessary information from collected data, filling in missing data, and standardizing it into text format.

[1251] A "generative artificial intelligence model" is a type of artificial intelligence that is trained using collected and pre-processed data to reproduce past behavioral patterns and decision-making of administrators.

[1252] An "emotion engine" is a device or software that includes technology to analyze facial expressions and voice in order to detect the user's emotional state.

[1253] An "inquiry" is the act of a user entering specific questions or requests for advice regarding a business.

[1254] "Processing result" refers to the response content generated based on the generative artificial intelligence model and the detected emotional state.

[1255] "Reflection" refers to the process by which a generative artificial intelligence model adjusts its response in consideration of the detected emotional state of the user.

[1256] This invention is a system that collects past behavioral data of corporate managers and constructs a virtual manager using a generative artificial intelligence model. Furthermore, by combining it with an emotion engine that recognizes the user's emotional state, it can provide more human-like responses. A specific embodiment of this system is described below.

[1257] Data Acquisition and Preprocessing

[1258] The server collects data from the company's database regarding past administrator presentations, speeches, and decision-making. This data undergoes preprocessing, including standardization to text format using regular expressions, removal of unnecessary information, and imputation of missing data. The preprocessed data is then used as training data for generative artificial intelligence models.

[1259] Training of generative AI models

[1260] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow and PyTorch. This training process analyzes past administrator behavior data to learn patterns of speech and decision-making styles.

[1261] Inquiry and Sentiment Recognition

[1262] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. These devices are equipped with an emotion engine that uses the camera and microphone to analyze the user's facial expressions and tone of voice, detecting their emotional state. For example, when a user inputs "Please tell me how to resolve the bottleneck in the production line," the emotion engine analyzes whether the user is tense or relaxed.

[1263] Query and sentiment data processing

[1264] The user's inquiry and emotional data are sent to the server in JSON format. The server analyzes the received data and inputs it into a generative artificial intelligence model. This model generates responses while taking the user's emotional state into consideration. For example, if the user is nervous, the response will be provided in a gentle tone.

[1265] Generating and providing answers

[1266] The AI ​​model's generated responses are validated by the server, and if appropriate, are sent directly to the user's device. The device then displays the received responses to the user. This entire process allows the user to receive appropriate advice from a virtual administrator. For example, specific responses such as, "To eliminate bottlenecks in the production line, it would be effective to review the worker shift schedule and utilize production management software," are provided.

[1267] Specific examples and prompt statements

[1268] As a concrete example, production reports from the past 10 years can be extracted from a factory's project logs. The user can then ask questions such as, "Please tell me specific ways to improve production efficiency," and the system will recognize their facial expressions using a camera and send the results to a server, providing appropriate advice. This allows the user to take concrete measures to improve productivity.

[1269] Example of a prompt

[1270] "Please tell me about specific measures to improve production efficiency."

[1271] "I'd like some advice on the timing of market launch."

[1272] "How can we optimize claims processing for a new production line?"

[1273] This invention makes it possible to provide appropriate decision-making support that takes into account the user's emotional state while utilizing the knowledge and experience of past administrators.

[1274] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1275] Step 1:

[1276] The server collects data from the company's database regarding past administrator presentations, speeches, and decisions. This involves a data collection process that executes database queries to extract the necessary information and convert it into an appropriate format. The input is the company's database, and the output is the collected raw data.

[1277] Step 2:

[1278] The server preprocesses the collected data. In this step, unnecessary information is removed using regular expressions, missing data is imputed, and the data is standardized to text format. This allows generative artificial intelligence models to efficiently learn from the data. The input is the collected raw data, and the output is the preprocessed clean data.

[1279] Step 3:

[1280] The server uses pre-processed data to train a generative artificial intelligence model using frameworks such as TensorFlow or PyTorch. During this process, the data is used as a training set for the model, learning the administrator's behavioral patterns and decision-making processes. The input is pre-processed clean data, and the output is the trained AI model.

[1281] Step 4:

[1282] Users input specific business inquiries and questions using tablet devices or smartphones within the factory. The user's input is presented as natural language prompts. The input is a text query from the user, and the output is data sent from the device to the server.

[1283] Step 5:

[1284] The emotion engine built into the device uses the camera and microphone to analyze the user's facial expressions and tone of voice to detect their emotional state. This emotion data is sent to the server in JSON format. The input is the user's voice and video data, and the output is the analyzed emotion data.

[1285] Step 6:

[1286] The server analyzes the received question content and sentiment data, and inputs it into a generative artificial intelligence model. The model considers the user's emotional state and generates an appropriate response. The input is the received question content and sentiment data, and the output is the response generated by the model.

[1287] Step 7:

[1288] The server validates the generated response and makes corrections if necessary. This validation includes checking the appropriateness and grammar of the response. The input is the generated response, and the output is the validated or corrected response.

[1289] Step 8:

[1290] The server sends the final answer to the terminal. The terminal displays the received answer to the user. This sequence of actions allows the user to receive appropriate advice from the virtual administrator. The input is the verified or corrected answer, and the output is the final answer presented to the user.

[1291] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 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.

[1292] 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.

[1293] 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 robot 414.

[1294] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1295] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1296] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1297] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1298] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1299] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1300] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1301] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1302] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1303] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1304] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1305] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1306] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1307] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1308] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1309] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1310] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1311] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[1312] The following is further disclosed regarding the embodiments described above.

[1313] (Claim 1)

[1314] A means of collecting data on the past behavior of managers within a company,

[1315] Means for preprocessing the collected data,

[1316] A means for training a generative artificial intelligence model using preprocessed data,

[1317] Means of receiving inquiries from users,

[1318] A means of processing received inquiries using a generative artificial intelligence model,

[1319] Means for providing the processing results to the user,

[1320] A system that includes this.

[1321] (Claim 2)

[1322] The system according to claim 1, characterized in that the generative artificial intelligence model utilizes natural language processing technology.

[1323] (Claim 3)

[1324] The system according to claim 1, characterized in that the administrator's behavioral data includes records of presentations, speeches, and decision-making.

[1325] "Example 1"

[1326] (Claim 1)

[1327] A means of collecting data on the past behavior of managers within a company,

[1328] A preprocessing method that standardizes the collected data into a text format, removes unnecessary information, and imparts missing data,

[1329] A means of training a generative AI model using preprocessed data,

[1330] A means of sending user inquiries to the server via the terminal,

[1331] A method for analyzing received inquiries, extracting key keywords, inputting them into a generation AI model, and generating answers,

[1332] The generated response is validated on a server, grammar checks and content appropriateness are verified, and adjustments are made as needed.

[1333] A means of sending the final answer from the server to the terminal and displaying it to the user,

[1334] A system that includes this.

[1335] (Claim 2)

[1336] The system according to claim 1, characterized in that the aforementioned generation AI model utilizes natural language processing technology.

[1337] (Claim 3)

[1338] The system according to claim 1, characterized in that the administrator's behavioral data includes records of presentations, speeches, and decision-making.

[1339] "Application Example 1"

[1340] (Claim 1)

[1341] A means of collecting data on the past behavior of managers within a company,

[1342] Means for preprocessing the collected data,

[1343] A means for training a generative artificial intelligence model using preprocessed data,

[1344] A means of collecting factory operation data in real time,

[1345] Means of receiving inquiries from users,

[1346] A means of processing received inquiries using a generative artificial intelligence model,

[1347] Means for providing the processing results to the user,

[1348] A system that includes this.

[1349] (Claim 2)

[1350] The system according to claim 1, characterized in that the generative artificial intelligence model utilizes natural language processing technology.

[1351] (Claim 3)

[1352] The system according to claim 1, characterized in that the administrator's behavioral data includes records of presentations, speeches, and decision-making.

[1353] (Claim 4)

[1354] The system according to claim 1, characterized by collecting and preprocessing data from sensors and cameras within a factory.

[1355] "Example 2 of combining an emotion engine"

[1356] (Claim 1)

[1357] A means of collecting data on the past behavior of managers within a company,

[1358] Means for preprocessing the collected data,

[1359] A means for training a generative artificial intelligence model using preprocessed data,

[1360] A means of receiving inquiries from users via a terminal,

[1361] A means of analyzing received inquiries and classifying them into appropriate themes and categories,

[1362] A means including an emotion engine that analyzes the user's emotional state through the terminal when an inquiry is made,

[1363] A means of inputting the analyzed emotional state and inquiry content into a generative artificial intelligence model for processing,

[1364] Means for verifying and adjusting the processing results,

[1365] Means for providing the processing results to the user via a terminal,

[1366] A system that includes this.

[1367] (Claim 2)

[1368] The system according to claim 1, characterized in that the generative artificial intelligence model utilizes natural language processing technology.

[1369] (Claim 3)

[1370] The system according to claim 1, characterized in that the administrator's behavioral data includes records of presentations, speeches, and decision-making.

[1371] "Application example 2 when combining with an emotional engine"

[1372] (Claim 1)

[1373] A means of collecting data on the past behavior of managers within a company,

[1374] Means for preprocessing the collected data,

[1375] A means for training a generative artificial intelligence model using preprocessed data,

[1376] Means of receiving inquiries from users,

[1377] A means of processing received inquiries using a generative artificial intelligence model,

[1378] Means for providing the processing results to the user,

[1379] A means for detecting a user's emotional state using an emotion engine,

[1380] A means of reflecting the user's emotional state in the response generation of a generative artificial intelligence model,

[1381] A system that includes this.

[1382] (Claim 2)

[1383] The system according to claim 1, characterized in that the generative artificial intelligence model utilizes natural language processing technology.

[1384] (Claim 3)

[1385] The system according to claim 1, characterized in that the administrator's behavioral data includes records of presentations, speeches, and decision-making.

[1386] (Claim 4)

[1387] The system according to claim 1, characterized in that the emotion engine analyzes the user's facial expressions and voice. [Explanation of Symbols]

[1388] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting data on the past behavior of managers within a company, Means for preprocessing the collected data, A means for training a generative artificial intelligence model using preprocessed data, Means of receiving inquiries from users, A means of processing received inquiries using a generative artificial intelligence model, Means for providing the processing results to the user, A system that includes this.

2. The system according to claim 1, characterized in that the generative artificial intelligence model utilizes natural language processing technology.

3. The system according to claim 1, characterized in that the administrator's behavioral data includes records of presentations, speeches, and decision-making.

Images