system

Abstract

We provide the system. [Solution] A data collection means for acquiring the user's past behavioral data and decision-making history, A data preprocessing means for cleansing and normalizing acquired data, A learning method that uses a deep learning model to learn the user's decision-making process, A recommendation generation means that generates recommendations tailored to the user's situation using a pre-trained model, A user feedback mechanism that obtains user feedback and sends it to the server, A system that includes model improvement methods for personalizing and improving the accuracy of models using user feedback and new data.

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The present invention relates to decision-making support for busy decision-makers such as business leaders, managers, and experts. In the conventional decision-making process, problems such as stress due to excessive information and lack of time have been issues, and there is a problem that it is difficult to make efficient and consistent decisions. Furthermore, current tools lack personalized support that reflects individual values, and the productivity of individuals and organizations cannot be maximized. Against this background, there is a demand for a system that internalizes individual values and decision-making processes and provides highly personalized support. <​​​​The present invention solves the above problems with a system comprising the following configuration: Data collection means acquires the user's past behavioral data and decision history. Data preprocessing means cleanses and normalizes the acquired data. Learning means learns the user's decision-making process using a deep learning model. Recommendation generation means generates recommendations tailored to the user's situation using the trained model. User feedback means acquires feedback from the user and sends it to the server. Model improvement means personalizes and improves the accuracy of the model using user feedback and new data. This makes it possible to optimize time management, reduce stress, and increase the productivity of individuals and organizations.

[0006] "Data collection means" refers to functions for acquiring users' past behavioral data and decision-making history.

[0007] "Data preprocessing means" refers to functions for cleansing and normalizing acquired data.

[0008] "Learning method" refers to a function that uses a deep learning model to learn the user's decision-making process.

[0009] "Recommendation generation means" refers to a function that uses a pre-trained model to generate recommendations tailored to the user's situation.

[0010] "User feedback means" refers to a function for obtaining feedback from users and sending it to the server.

[0011] "Model improvement measures" refer to functions that personalize and improve the accuracy of a model using user feedback and new data. [Brief explanation of the drawing]

[0012] [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 the data processing device and 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] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention]

[0013] Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings.

[0014] First, the terms used in the following description will be explained.

[0015] In the following embodiments, a processor with a reference numeral (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.

[0016] In the following embodiments, a RAM (Random Access Memory) with a reference numeral is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0017] In the following embodiments, a storage with a reference numeral is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, etc.

[0018] In the following embodiments, a communication I / F (Interface) with a reference numeral is an interface that includes a communication processor and an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark), etc.

[0019] 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."

[0020] [First Embodiment]

[0021] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0022] 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.

[0023] 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).

[0024] 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.

[0025] 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.

[0026] 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.

[0027] 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.

[0028] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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".

[0033] This invention is a system that supports decision-making for busy business leaders, managers, and experts, and its specific form is shown below.

[0034] System Configuration

[0035] The system of this invention mainly consists of the following elements:

[0036] 1. Data acquisition methods

[0037] 2. Data preprocessing means

[0038] 3. Learning methods

[0039] 4. Recommended generation method

[0040] 5. User Feedback Methods

[0041] 6. Model Improvement Methods

[0042] System operation

[0043] Data acquisition methods

[0044] The server retrieves the user's past behavioral data and decision-making history from a database. This includes purchase history, schedules, and message exchanges.

[0045] The terminal monitors user operations and actions in real time and sends this information to the server.

[0046] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This information is also sent to the server and stored in the database.

[0047] Data preprocessing means

[0048] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analysis.

[0049] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[0050] Learning methods

[0051] The server learns the user's decision-making process using deep learning models (e.g., recurrent neural networks or transformer models). This allows it to extract individual user patterns and incorporate them into the model.

[0052] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[0053] Recommended generation method

[0054] The server uses a pre-trained model to generate user-specific recommendations. For example, when a user is deciding on project resource allocation, it suggests the best option based on past success rates and cost-effectiveness.

[0055] The terminal notifies the user of recommendations from the server and displays them on the interface.

[0056] User Feedback Methods

[0057] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, they also send the results to the server.

[0058] The device collects user feedback in real time and sends it to the server.

[0059] Model Improvement Means

[0060] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[0061] Specific example

[0062] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0063] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[0064] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[0065] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[0066] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[0067] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[0068] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0069] In this way, the system of the present invention supports the user's decision-making process and provides optimized decision-making. This enables personalized support based on individual values, improving the user's productivity and efficiency.

[0070] The following describes the processing flow.

[0071] ---

[0072] Step 1:

[0073] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[0074] The device monitors the user's real-time operations and actions and sends that data to the server.

[0075] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server.

[0076] Step 2:

[0077] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[0078] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[0079] Step 3:

[0080] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data and values. This allows it to model the user's decision-making process.

[0081] The trained model reflects the user's characteristics and past decision-making patterns.

[0082] Step 4:

[0083] The server uses a pre-trained model to generate recommendations based on the user's current situation. For example, it might recommend which project the user should allocate resources to next.

[0084] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[0085] Step 5:

[0086] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[0087] The device collects user feedback in real time and sends that data to the server.

[0088] Step 6:

[0089] The server uses user feedback and new behavioral data to retrain the model, aiming for personalization and improved accuracy.

[0090] This will improve the quality of future recommendations and more effectively support user decision-making.

[0091] ---

[0092] The above describes the specific processing flow in the system of the present invention. This enables efficient and consistent support for the user's decision-making process.

[0093] (Example 1)

[0094] 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."

[0095] Traditional decision support systems have faced challenges in effectively assisting busy business leaders and managers with their decision-making. In particular, their inability to provide personalized recommendations that reflect individual user values ​​and priorities often leads to a decline in recommendation quality and reduced decision-making efficiency. Furthermore, the difficulty in effectively incorporating user feedback to improve model accuracy results in a gradual deterioration of recommendation quality.

[0096] 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.

[0097] In this invention, the server includes data collection means for acquiring the user's past behavioral data and decision-making history, data preprocessing means for cleansing and normalizing the acquired data, learning means using a deep learning model, recommendation generation means for generating recommendations, user feedback means for receiving user feedback, and model improvement means for improving the accuracy of the model. This makes it possible to provide personalized recommendations based on the user's individual behavior and values, and to continuously improve the quality of recommendations and the efficiency of decision-making.

[0098] "Data collection methods" refer to means of acquiring users' past behavioral data and decision-making history.

[0099] "Data preprocessing means" refers to methods for cleansing and normalizing acquired data.

[0100] A "learning method" is a means of learning user decision-making patterns using a deep learning model.

[0101] A "recommendation generation method" is a method for generating recommendations for users using a pre-trained model.

[0102] A "user feedback method" is a means of receiving user feedback.

[0103] "Model improvement methods" are means of improving the accuracy of a model based on user feedback.

[0104] This invention is a system that supports decision-making for busy business leaders, managers, and professionals. This system primarily consists of the following elements:

[0105] 1. Data acquisition methods

[0106] 2. Data preprocessing means

[0107] 3. Learning methods

[0108] 4. Recommended generation method

[0109] 5. User Feedback Methods

[0110] 6. Model Improvement Methods

[0111] Data acquisition methods

[0112] The server retrieves data on the user's past behavior and decision-making history from the database. For example, it collects data such as purchase history, schedules, and message exchanges. This provides the foundational data needed to understand the user's decision-making patterns.

[0113] The device monitors user actions and behaviors in real time and sends that information to the server. For example, it monitors actions such as a user adding an event to their calendar in real time and sends that data to the server.

[0114] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[0115] Data preprocessing means

[0116] The server cleanses the acquired data, removing inconsistent data. Furthermore, it normalizes the data and converts it into a format suitable for analysis. For example, it might unify date data in different formats.

[0117] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. This process ensures data consistency.

[0118] Learning methods

[0119] The server learns user decision-making patterns using deep learning models (e.g., recurrent neural networks or transformer models). Specifically, data is divided into batches and fed into the model, and learning is performed epoch by epoch.

[0120] The training process is periodically rerun in batches, and the model is continuously updated with new data. This process allows the model to make recommendations that reflect the latest data.

[0121] Recommended generation method

[0122] The server uses a pre-trained model to generate personalized recommendations based on the user's situation. For example, when a business leader is deciding which projects to allocate resources to next month, the server suggests the best options based on past success rates and cost-effectiveness.

[0123] The terminal notifies the user of recommendations from the server and displays them on the interface. Notifications are made in real time, allowing the user to check them immediately.

[0124] User Feedback Methods

[0125] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, the results are also sent to the server.

[0126] The device collects user feedback in real time and sends it to the server. The collected data is immediately reflected in the database.

[0127] Model Improvement Means

[0128] The server improves the model's accuracy using user feedback and newly collected data. By analyzing the feedback data and optimizing the model's loss function, the accuracy of the next recommendation improves.

[0129] Specific examples and prompt statements

[0130] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0131] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[0132] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[0133] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[0134] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[0135] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[0136] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0137] Examples of prompts for a generative AI model:

[0138] "Based on past project data, please recommend the optimal resource allocation for the next month. Consider variables such as success rate, cost, and time."

[0139] By using the above system, it becomes possible to support the user's decision-making process and provide personalized recommendations, thereby improving productivity and efficiency.

[0140] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0141] Step 1: Data Collection

[0142] The server retrieves the user's past behavioral data and decision-making history from the database. Specifically, it issues queries to collect data such as purchase history, schedules, and message exchanges. The input is the query results from the database, and the output is the set of collected data.

[0143] The device monitors user actions and behaviors in real time and sends that information to the server. Specifically, it captures actions such as a user adding an event to their calendar and sends that data to the server. The input is the user's real-time actions, and the output is the monitored data.

[0144] As part of the initial setup in the form of a questionnaire, users enter their values ​​and priorities. This data is also sent to the server and stored in the database. The input is the user's questionnaire responses, and the output is the data sent to the server.

[0145] Step 2: Data Preprocessing

[0146] The server cleanses the acquired data and removes inconsistent data. Specifically, it imputes missing values ​​and removes outliers. The input is the collected raw data, and the output is the cleansed data.

[0147] The server further normalizes the data and standardizes its format. For example, it unifies date data in different formats. The input is cleansed data, and the output is normalized data.

[0148] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. The input is the user's survey data, and the output is the converted data.

[0149] Step 3: Model Training

[0150] The server trains a deep learning model using normalized data. Specifically, it batch processes the data using recurrent neural networks or transformer models and performs training in epoch units. The input is a normalized dataset, and the output is the trained model.

[0151] The training process is periodically rerun in batches, and the model is continuously updated using newly collected data. The input is the new data, and the output is the updated model.

[0152] Step 4: Recommended Generation

[0153] The server uses a trained model to generate user-specific recommendations. For example, it might suggest the optimal project resource allocation for a business leader for the following month. The input is the trained model and current user data, and the output is the generated recommendations.

[0154] The terminal notifies the user of recommendations received from the server and displays them on the interface. Specifically, it presents recommendations to the user through pop-up notifications. The input is the recommendations, and the output is the notification to the user.

[0155] Step 5: User Feedback

[0156] Users review the provided recommendations and make corrections or provide feedback as needed. For example, they may be given options to accept, reject, or modify the recommendations. The input is the user's feedback, and the output is the modified or accepted data.

[0157] The device collects user feedback in real time and sends it to the server. The input is the user's real-time feedback, and the output is the data sent to the server.

[0158] Step 6: Model Improvement

[0159] The server improves the model's accuracy using user feedback and newly collected data. Specifically, it analyzes the feedback data and optimizes the model's loss function. The input is the feedback data and newly collected data, and the output is the updated model.

[0160] By following these steps, the system can continue to provide recommendations optimized for the user's situation.

[0161] (Application Example 1)

[0162] 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."

[0163] In conventional factory robot operations, it was difficult to efficiently manage the operating data, maintenance records, and production plans of individual robots, making it challenging to determine the optimal operating schedule and maintenance timing. As a result, production efficiency decreased and costs increased due to unnecessary maintenance. The present invention aims to solve these problems and provide a new system for optimizing robot operations within a factory.

[0164] 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.

[0165] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, means for acquiring and analyzing data to optimize the operational efficiency of robots in a factory, means for recommending an optimal operational schedule and maintenance timing, and means for displaying the recommended operational schedule and maintenance timing. This enables efficient operational management and appropriate maintenance of factory robots.

[0166] "Data collection means" refers to means for acquiring information such as users' past behavioral data, decision-making history, robot operation data within the factory, maintenance records, and production plans.

[0167] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, and further converting it into a format suitable for analyzing the operational efficiency of robots within a factory.

[0168] "Learning methods" refer to techniques that improve the accuracy of recommendations by using deep learning models based on user and robot behavior data.

[0169] A "recommendation generation method" is a means of generating recommendations to optimize the next task or maintenance timing of a factory robot using a pre-trained model.

[0170] A "user feedback mechanism" is a means of collecting user feedback and modifications to generated recommendations and incorporating them into the system.

[0171] "Model improvement methods" refer to techniques for continuously improving the accuracy of a learning model using user feedback and new data.

[0172] "Means of acquiring and analyzing data to optimize operational efficiency" refers to methods of collecting data on robots within a factory and analyzing it to achieve optimal operational efficiency.

[0173] "Means for recommending optimal operating schedules and maintenance timings" refers to methods for proposing optimal operating schedules and maintenance timings for robots to users based on collected and analyzed data.

[0174] "Means for displaying recommended operational schedules and maintenance timings" refers to a means of visually communicating the generated recommendations by displaying them on the user's device.

[0175] This invention is a system for optimizing the operational efficiency of robots in a factory, and its specific form is shown below.

[0176] System Configuration

[0177] The system of the present invention consists of the following elements:

[0178] 1. Data acquisition methods

[0179] 2. Data preprocessing means

[0180] 3. Learning methods

[0181] 4. Recommended generation method

[0182] 5. User Feedback Methods

[0183] 6. Model Improvement Methods

[0184] 7. Means for acquiring and analyzing data to optimize operational efficiency

[0185] 8. Means for recommending optimal operating schedules and maintenance timings.

[0186] 9. Means for displaying recommended operational schedules and maintenance timings.

[0187] System operation

[0188] Data acquisition methods

[0189] The server retrieves operational data, maintenance records, and production plans from robots within the factory via an API.

[0190] Users provide past behavioral data and decision-making history, which is also sent to the server and stored in the database.

[0191] Data preprocessing means

[0192] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analyzing operational efficiency within the factory.

[0193] The terminal converts the data obtained from the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[0194] Learning methods

[0195] The server uses deep learning models (e.g., recurrent neural networks or transformer models) to learn user and robot behavior data and incorporate it into the model.

[0196] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[0197] Recommended generation method

[0198] The server uses pre-trained models to recommend optimizations for the next tasks and maintenance timings of factory robots.

[0199] The terminal notifies the user of recommendations from the server and displays them on the interface.

[0200] User Feedback Methods

[0201] Users review the provided recommendations and make corrections or provide feedback. This feedback is sent to the server.

[0202] The device collects user feedback in real time and sends it to the server.

[0203] Model Improvement Means

[0204] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[0205] Specific example

[0206] As a concrete example, consider a scenario where a factory manager decides which robots should be assigned which tasks in the following month.

[0207] 1. Data collection method: The server acquires historical data such as the operation data, success rate, maintenance records, and production plans of factory robots.

[0208] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the data into a format for analyzing the robot's operational efficiency and sends it to the server.

[0209] 3. Learning method: The server uses a deep learning model (e.g., Keras / TensorFlow) to learn the robot's operational patterns.

[0210] 4. Recommendation generation method: The server generates recommendations for the optimal task assignment for the following month and notifies the manager via the terminal.

[0211] 5. User Feedback Method: Managers review, approve, or revise recommendations. Feedback is sent to the server via the terminal.

[0212] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0213] This optimizes the operational schedule and maintenance timing of robots within the factory, improving productivity and efficiency.

[0214] Example of a prompt

[0215] "Please retrieve robot operation data from the factory data API."

[0216] Please normalize your data using Pandas and Scikit-learn.

[0217] "Load the saved Keras model and perform inference."

[0218] "Generate recommendations based on the model's inference results."

[0219] "Please display the generated recommendations in JSON format."

[0220] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0221] Step 1:

[0222] The server retrieves information such as robot operation data, maintenance records, and production plans from robots within the factory via an API. This includes each robot's work history, current status, and future work schedule. Input is raw data retrieved from the API, and output is canonical data stored in the database.

[0223] Step 2:

[0224] The server cleanses the acquired raw data and removes inconsistent data. Next, it normalizes the data and converts it into a format suitable for analyzing the operational efficiency of factory robots. Data processing libraries such as Pandas and Scikit-learn are used for this. The input is the raw data acquired in step 1, and the output is the cleansed and normalized data.

[0225] Step 3:

[0226] The server uses a deep learning model (e.g., Keras / TensorFlow) to train on normalized data. The training process involves pattern extraction and model training using recurrent neural networks or transformer models based on acquired historical data. The input is the data preprocessed in step 2, and the output is the trained model.

[0227] Step 4:

[0228] The server uses a pre-trained model to perform inference and recommend optimizations for the next task and maintenance timing of factory robots. This involves converting the generated recommendations into JSON format and sending them to the terminal. The input is the pre-trained model and current data, and the output is the recommendations in JSON format.

[0229] Step 5:

[0230] The terminal notifies the user of recommendations received from the server and displays them on the interface. This allows the user to review the recommendations and provide feedback or make corrections as needed. The input is the recommendations received from the server in JSON format, and the output is the recommendations displayed in a format usable by the user.

[0231] Step 6:

[0232] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The feedback is sent to the server via the device. The input is the recommendations displayed on the device, and the output is the feedback sent to the server.

[0233] Step 7:

[0234] The server uses received user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided. The input is user feedback and new data, and the output is the updated trained model.

[0235] 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.

[0236] This invention is a system that supports decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes the user's emotions. Its specific form is described below.

[0237] System Configuration

[0238] The system of the present invention consists of the following elements:

[0239] 1. Data acquisition methods

[0240] 2. Data preprocessing means

[0241] 3. Learning methods

[0242] 4. Recommended generation method

[0243] 5. User Feedback Methods

[0244] 6. Model Improvement Methods

[0245] 7. Emotional Engine

[0246] System operation

[0247] Data acquisition methods

[0248] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[0249] The device monitors the user's real-time operations and actions and sends that data to the server.

[0250] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[0251] The emotion engine acquires emotion data from the user's voice, facial expressions, text, etc., and sends it to the server.

[0252] Data preprocessing means

[0253] The server cleanses all acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[0254] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[0255] The emotion data acquired by the emotion engine is also appropriately preprocessed before being sent to the server.

[0256] Learning methods

[0257] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process.

[0258] The trained model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[0259] Recommended generation method

[0260] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it might recommend which project to allocate resources to next, taking into account whether the user is feeling uplifted or stressed.

[0261] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[0262] User Feedback Methods

[0263] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide feedback regarding their own feelings.

[0264] The device collects user feedback in real time and sends that data to the server.

[0265] Model Improvement Means

[0266] The server uses user feedback, new behavioral data, and sentiment data to retrain the model, aiming for personalization and improved accuracy.

[0267] This will improve the quality of future recommendations and more effectively support user decision-making.

[0268] Specific example

[0269] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0270] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data. The emotion engine collects emotional data from the leader's voice and facial expressions.

[0271] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[0272] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[0273] 4. Recommended generation method: The server recommends high-risk projects when the user is feeling exhilarated, and stable projects when the user is feeling stressed.

[0274] 5. User Feedback Methods: Leaders review, approve, or revise recommendations. They also provide emotional feedback.

[0275] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0276] This enables the system of the present invention to support the user's decision-making process and provide personalized recommendations that take emotions into consideration. This, in turn, improves user productivity and efficiency and contributes to stress reduction.

[0277] The following describes the processing flow.

[0278] ---

[0279] Step 1:

[0280] The server retrieves the user's past behavior data and decision-making history from the database. This includes purchase history, project evaluations, results of past decisions, and so on.

[0281] The terminal monitors the user's real-time operations and actions and sends that data to the server. For example, which web pages the user is viewing, which apps the user is using, and so on.

[0282] The emotion engine obtains emotion data from the user's voice, facial expressions, and text. For example, it analyzes the stress level and happiness from the tone of the user's speech and facial expressions.

[0283] Step 2:

[0284] The server cleans all the acquired data and removes inaccurate data and noise. For example, it removes duplicate data and inconsistent data.

[0285] The server performs normalization and converts the data into a form suitable for analysis. For example, it scales numerical data or encodes categorical data.

[0286] The terminal converts the questionnaire results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server. For example, it extracts text data from the questionnaire form and converts it into JSON format. <​​​​​​​​​​The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process.

[0290] The training process is periodically rerun in batches, and the model is continuously updated with new data. For example, the model is retrained every night with new data.

[0291] Step 4:

[0292] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is feeling stressed, it will recommend safer options.

[0293] The device notifies the user of recommendations received from the server and displays them on the interface. For example, it might use smartphone push notifications or notifications from wearable devices.

[0294] Step 5:

[0295] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[0296] The device collects user feedback in real time and sends that data to the server. For example, when a user presses the "Approve" button in a push notification, that selection is recorded and sent.

[0297] Step 6:

[0298] The server uses user feedback and new behavioral and sentiment data to retrain its model, aiming for personalization and improved accuracy.

[0299] As a result, the quality of subsequent recommendations can be improved, and it becomes possible to more effectively support the user's decision-making. For example, new recommendations are re-adjusted based on the recommendations the user accepted last time and their subsequent emotional reactions.

[0300] ---

[0301] The above is the specific flow of processing in the system of the present invention combined with the emotion engine. As a result, it is realized that the user's decision-making process is efficiently and consistently supported while taking emotions into account.

[0302] (Example 2)

[0303] Next, Example 2 will be described. In the following description, the data processing device 12 is referred to as a "server", and the smart device 14 is referred to as a "terminal".

[0304] There is a problem that busy business leaders, managers, experts, etc. are easily affected by emotions when making decisions quickly and accurately. Also, this causes problems such as stress and reduced productivity. Therefore, there is a need for a system that provides personalized recommendations while considering the user's emotional state and supports decision-making.

[0305] The specific processing by the specific processing unit 290 of the data processing device 12 in Example 2 is realized by the following respective means.

[0306] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and emotion data acquisition means. As a result, it becomes possible to provide personalized and emotion-responsive recommendations based on the user's behavior data and emotion data. [[ID=3o]]

[0307] The "data collection means" is means for acquiring the user's past behavior data and decision-making history and transmitting them to the server.

[0308] "Data preprocessing means" refers to means for cleansing and normalizing acquired data and converting it into a format suitable for analysis.

[0309] "Learning methods" refer to techniques for performing deep learning using recurrent neural networks (RNNs) or transformer models based on user behavior data, values, and emotional data.

[0310] A "recommendation generation method" is a means of generating recommendations based on the current user's situation and emotions, using a pre-trained model.

[0311] A "user feedback mechanism" is a means for users to provide feedback on the recommendations they have been given, to collect that feedback in real time, and to send it to the server.

[0312] "Model improvement methods" refer to techniques for retraining a model using user feedback, new behavioral data, and sentiment data, with the aim of improving personalization and accuracy.

[0313] "Means for acquiring emotional data" refers to a means of collecting emotional data from the user's voice, facial expressions, and text, and transmitting it to a server.

[0314] This invention provides a system to support decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes user emotions. The system of this invention includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and emotion data acquisition means. Its specific form is shown below.

[0315] 1. Data acquisition methods

[0316] The server retrieves the user's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. For example, the server uses an API to collect this data.

[0317] The device monitors the user's real-time actions and behaviors, including the location and frequency of clicks and taps, and sends this data to a server. This is done using the device's sensors.

[0318] As part of the initial setup, users enter their values ​​and priorities into the system via web forms or in-app questionnaires. This input data is sent to the server via the device and stored in a database.

[0319] The emotion engine acquires emotional data from the user's voice, facial expressions, text, etc., and sends it to the server. For example, it performs facial recognition through the camera and voice analysis through the microphone.

[0320] 2. Data preprocessing means

[0321] The server cleanses all acquired data, removing inaccuracies and noise. Specifically, it removes duplicate data and imputes missing values. It also normalizes the data, converting data from different data sources into a unified format.

[0322] The terminal converts the survey results entered by the user into an appropriate data format, such as JSON or XML, and sends that data to the server.

[0323] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[0324] 3. Learning methods

[0325] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The model is trained on a training dataset, and its parameters are adjusted. The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[0326] 4. Recommended generation method

[0327] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[0328] The device displays recommendations received from the server to the user via an interface. Information is provided in real time using notifications and pop-up messages.

[0329] 5. User Feedback Methods

[0330] Users can review the provided recommendations and approve, reject, or modify them. They can also provide feedback on whether the recommendations were appropriate and whether they fit their emotional state.

[0331] The device collects user feedback in real time and sends the results to the server.

[0332] 6. Model Improvement Methods

[0333] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[0334] Specific example

[0335] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0336] The server retrieves historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[0337] The emotion engine collects emotional data from the leader's voice and facial expressions.

[0338] The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[0339] The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[0340] The server recommends high-risk projects when you're feeling exhilarated, and stable projects when you're feeling stressed.

[0341] Leaders review, approve, or revise recommendations, and also provide emotional feedback.

[0342] The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0343] This enables the system of the present invention to support user decision-making and provide personalized recommendations that take emotions into consideration. Furthermore, this improves user productivity and efficiency and contributes to stress reduction.

[0344] Example of a prompt

[0345] The system generates an answer to the question, "Which project should we allocate resources to next month?" by considering the success rate of recent projects, resource utilization, and leader emotional data (emotional state, stress level, etc.).

[0346] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0347] Step 1:

[0348] Data collection:

[0349] The server retrieves users' past behavioral data and decision-making history from the database using multiple APIs. This includes purchase history, schedules, and message exchanges.

[0350] Input: User ID

[0351] Output: User data such as purchase history, schedule, and message history.

[0352] The device monitors the user's real-time actions (location and frequency of clicks and taps) and behavior, and periodically sends this data to the server. Activity data is captured using the device's sensors.

[0353] Input: User operation data

[0354] Output: Real-time operation data

[0355] As part of the initial setup, users enter their values ​​and priorities through web forms or in-app questionnaires. This data is transmitted to the server via the device and stored in a database.

[0356] Input: Survey data

[0357] Output: Values ​​and priority data

[0358] The emotion engine acquires emotional data from the user's voice, facial expressions, and text, and sends it to the server. For example, it recognizes facial expressions through the camera and performs voice analysis through the microphone.

[0359] Input: Voice, facial expressions, text data

[0360] Output: Sentiment data

[0361] Step 2:

[0362] Data preprocessing:

[0363] The server cleanses the acquired data, removing inaccuracies and noise. Specific tasks include deleting duplicate data and imputing missing values. It also normalizes the data, converting data from different sources into a unified format.

[0364] Input: Raw data (purchase history, schedule, message history, real-time operation data, sentiment data)

[0365] Output: Cleansed and normalized data

[0366] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[0367] Input: Survey results

[0368] Output: Survey data in JSON or XML format

[0369] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[0370] Input: Sentiment data

[0371] Output: Preprocessed sentiment data

[0372] Step 3:

[0373] Model learning:

[0374] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process. The model is trained on a training dataset and its parameters are adjusted.

[0375] Input: Cleansed and normalized data

[0376] Output: Trained model

[0377] The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[0378] Input: Training data

[0379] Output: High-precision model

[0380] Step 4:

[0381] Recommended generation:

[0382] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[0383] Input: Trained model, current situation data, sentiment data

[0384] Output: Recommendations

[0385] The terminal notifies the user of recommendations received from the server and displays them on the interface. Information is provided in real time using notifications and pop-up messages.

[0386] Input: Recommendations

[0387] Output: Display on the user screen

[0388] Step 5:

[0389] Collecting user feedback:

[0390] Users review the provided recommendations and can approve, reject, or modify them. They also provide feedback on whether the recommendations are appropriate and whether they align with their emotional state.

[0391] Input: User feedback on recommendations

[0392] Output: Revised recommendations, feedback data

[0393] The device collects user feedback in real time and sends the results to the server.

[0394] Input: User feedback data

[0395] Output: Feedback data sent to the server

[0396] Step 6:

[0397] Model improvements:

[0398] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[0399] Input: Feedback data, new behavioral data, emotional data

[0400] Output: Improved model

[0401] This allows the system to support user decision-making, provide personalized recommendations that take emotions into account, and improve user productivity and efficiency, as well as reduce stress.

[0402] (Application Example 2)

[0403] 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".

[0404] In modern factories and production sites, decision-making regarding robot operation and maintenance is crucial. However, many systems fail to consider the emotions and stress levels of managers, leading to inefficient resource allocation and increased robot failure risks. Therefore, decision support systems that take managers' emotions into account need to provide more optimal resource allocation and maintenance schedules.

[0405] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means. This makes it possible to grasp the emotional state of the administrator in real time and recommend appropriate resource allocation and maintenance schedules based on that.

[0406] "Data collection means" refers to methods for acquiring users' past behavioral data and decision-making history, as well as real-time operation data and emotional data.

[0407] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, removing inaccurate data and noise, and converting it into a format suitable for analysis.

[0408] "Learning methods" refer to techniques that use deep learning models or machine learning algorithms to model user decision-making patterns based on user behavior data, decision history, values ​​data, sentiment data, etc.

[0409] A "recommendation generation method" is a means of generating appropriate recommendations based on the current user situation and sentiment data, using a trained model.

[0410] A "user feedback mechanism" refers to a means of collecting feedback from users, such as modifications or approvals of recommendations provided by users.

[0411] "Model improvement methods" refer to techniques for retraining a model using user feedback and new behavioral and sentiment data to improve the accuracy of recommendations.

[0412] "Emotion recognition means" refers to methods for acquiring and analyzing emotional data from a user's voice, facial expressions, text, etc.

[0413] A "real-time decision support system" is a means of providing appropriate recommendations immediately based on data and sentiment data collected in real time, thereby supporting the user's decision-making.

[0414] The embodiments for carrying out this invention will be described below. The factory robot management assistant system includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means.

[0415] System Configuration

[0416] Data acquisition methods

[0417] In this system, the server retrieves the administrator's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. The terminal monitors real-time operation data and sends it to the server. The emotion engine, used as an emotion recognition tool, retrieves emotion data from the administrator's voice, facial expressions, text, etc., and sends it to the server.

[0418] Data preprocessing means

[0419] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis. On the terminal side, the survey results entered by the user are converted into an appropriate data format (e.g., JSON or XML) and sent to the server. Sentiment data is also preprocessed and sent to the server.

[0420] Learning methods

[0421] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The learned model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[0422] Recommended generation method

[0423] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it recommends high-risk projects when the administrator is emotionally uplifted, and stable projects when they are stressed. The terminal notifies the user of the recommendations received from the server and displays them in the interface.

[0424] User Feedback Methods

[0425] Users review the provided recommendations and provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide emotional feedback. The device collects user feedback in real time and sends the data to the server.

[0426] Model Improvement Means

[0427] The server retrains its model using user feedback and new behavioral and sentiment data to improve the accuracy of its recommendations.

[0428] Hardware and software to be used

[0429] Hardware: Computer, camera, and microphone built into the factory robot.

[0430] Software: Python, scikit-learn, emotion_recognition library

[0431] Processing Overview

[0432] 1. Data collection:

[0433] The server retrieves the administrator's past behavioral data and decision-making history from a database, while the terminal monitors real-time operation data. The emotion engine collects emotional data from the administrator's voice and facial expressions.

[0434] 2. Data preprocessing:

[0435] The acquired data is cleansed and normalized by the server. Data such as survey results are also converted into an appropriate format.

[0436] 3. Learning:

[0437] The server uses a deep learning model to learn from the administrator's behavioral and emotional data.

[0438] 4. Recommended generation:

[0439] Based on real-time data and pre-trained models, it generates recommendations tailored to the current administrator situation.

[0440] 5. Gathering feedback:

[0441] User feedback is sent to the server via the device.

[0442] 6. Model Improvement:

[0443] The server retrains the model based on the feedback, improving the accuracy of future recommendations.

[0444] Specific example

[0445] For example, if a factory robot manager is under a very high level of stress, this management assistant app will recommend reducing the high-risk tasks performed by the robots. Conversely, if the manager is emotionally stable, it will prioritize efficiency and make recommendations that include high-risk tasks.

[0446] Example of a prompt

[0447] "Based on the following data, please generate the optimal work schedule for factory robot management. The data is as follows:"

[0448] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0449] Step 1:

[0450] The server retrieves the administrator's past behavioral data and decision history from the database. This includes information such as purchase history, schedules, and message exchanges. The input is past behavioral data and decision history from the database, and the output is a dataset containing this data. Specifically, it uses SQL queries to retrieve the necessary information from the database.

[0451] Step 2:

[0452] The terminal monitors real-time operation and emotion data and sends it to the server. Inputs are administrator operations and voice and facial expression data collected by the emotion engine, while output is real-time data sent to the server. Specifically, emotion recognition software is used to extract emotion information from voice and facial expressions.

[0453] Step 3:

[0454] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data. The input is the data obtained in steps 1 and 2, and the output is the cleansed and normalized dataset. Specifically, it uses a data cleaning tool to remove incomplete data and noise, and a standard scaler to normalize the data.

[0455] Step 4:

[0456] The server performs deep learning using recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and sentiment data. The input is a cleansed and normalized dataset, and the output is the trained recommendation model. Specifically, the model is trained using deep learning frameworks such as PyTorch and TensorFlow.

[0457] Step 5:

[0458] The server uses a pre-trained model to generate recommendations based on the current user's situation and sentiment. The input is real-time interaction and sentiment data, and the output is recommendations. Specifically, it inputs real-time data into the pre-trained model and generates recommendations.

[0459] Step 6:

[0460] The terminal notifies the user of recommendations received from the server and displays them on the interface. The input is the recommendations from the server, and the output is the user's feedback. Specifically, it displays the recommendations using the user interface (UI) and receives feedback from the user.

[0461] Step 7:

[0462] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The input is the recommendations displayed on the device, and the output is the feedback sent to the device. Specifically, the user approves, rejects, or modifies the recommendations and inputs the result into the device.

[0463] Step 8:

[0464] The server retrains the model using user feedback and new behavioral and sentiment data. The input is user feedback data and new data, and the output is the updated trained model. Specifically, it retrains the model using feedback data to improve accuracy.

[0465] 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.

[0466] 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.

[0467] 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.

[0468] [Second Embodiment]

[0469] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0470] 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.

[0471] 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).

[0472] 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.

[0473] 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.

[0474] 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).

[0475] 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.

[0476] 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.

[0477] 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.

[0478] 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.

[0479] 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.

[0480] 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".

[0481] This invention is a system that supports decision-making for busy business leaders, managers, and experts, and its specific form is shown below.

[0482] System Configuration

[0483] The system of this invention mainly consists of the following elements:

[0484] 1. Data acquisition methods

[0485] 2. Data preprocessing means

[0486] 3. Learning methods

[0487] 4. Recommended generation method

[0488] 5. User Feedback Methods

[0489] 6. Model Improvement Methods

[0490] System operation

[0491] Data acquisition methods

[0492] The server retrieves the user's past behavioral data and decision-making history from a database. This includes purchase history, schedules, and message exchanges.

[0493] The terminal monitors user operations and actions in real time and sends this information to the server.

[0494] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This information is also sent to the server and stored in the database.

[0495] Data preprocessing means

[0496] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analysis.

[0497] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[0498] Learning methods

[0499] The server learns the user's decision-making process using deep learning models (e.g., recurrent neural networks or transformer models). This allows it to extract individual user patterns and incorporate them into the model.

[0500] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[0501] Recommended generation method

[0502] The server uses a pre-trained model to generate user-specific recommendations. For example, when a user is deciding on project resource allocation, it suggests the best option based on past success rates and cost-effectiveness.

[0503] The terminal notifies the user of recommendations from the server and displays them on the interface.

[0504] User Feedback Methods

[0505] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, they also send the results to the server.

[0506] The device collects user feedback in real time and sends it to the server.

[0507] Model Improvement Means

[0508] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[0509] Specific example

[0510] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0511] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[0512] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[0513] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[0514] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[0515] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[0516] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0517] In this way, the system of the present invention supports the user's decision-making process and provides optimized decision-making. This enables personalized support based on individual values, improving the user's productivity and efficiency.

[0518] The following describes the processing flow.

[0519] ---

[0520] Step 1:

[0521] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[0522] The device monitors the user's real-time operations and actions and sends that data to the server.

[0523] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server.

[0524] Step 2:

[0525] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[0526] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[0527] Step 3:

[0528] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data and values. This allows it to model the user's decision-making process.

[0529] The trained model reflects the user's characteristics and past decision-making patterns.

[0530] Step 4:

[0531] The server uses a pre-trained model to generate recommendations based on the user's current situation. For example, it might recommend which project the user should allocate resources to next.

[0532] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[0533] Step 5:

[0534] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[0535] The device collects user feedback in real time and sends that data to the server.

[0536] Step 6:

[0537] The server uses user feedback and new behavioral data to retrain the model, aiming for personalization and improved accuracy.

[0538] This will improve the quality of future recommendations and more effectively support user decision-making.

[0539] ---

[0540] The above describes the specific processing flow in the system of the present invention. This enables efficient and consistent support for the user's decision-making process.

[0541] (Example 1)

[0542] 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".

[0543] Traditional decision support systems have faced challenges in effectively assisting busy business leaders and managers with their decision-making. In particular, their inability to provide personalized recommendations that reflect individual user values ​​and priorities often leads to a decline in recommendation quality and reduced decision-making efficiency. Furthermore, the difficulty in effectively incorporating user feedback to improve model accuracy results in a gradual deterioration of recommendation quality.

[0544] 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.

[0545] In this invention, the server includes data collection means for acquiring the user's past behavioral data and decision-making history, data preprocessing means for cleansing and normalizing the acquired data, learning means using a deep learning model, recommendation generation means for generating recommendations, user feedback means for receiving user feedback, and model improvement means for improving the accuracy of the model. This makes it possible to provide personalized recommendations based on the user's individual behavior and values, and to continuously improve the quality of recommendations and the efficiency of decision-making.

[0546] "Data collection methods" refer to means of acquiring users' past behavioral data and decision-making history.

[0547] "Data preprocessing means" refers to methods for cleansing and normalizing acquired data.

[0548] A "learning method" is a means of learning user decision-making patterns using a deep learning model.

[0549] A "recommendation generation method" is a method for generating recommendations for users using a pre-trained model.

[0550] A "user feedback method" is a means of receiving user feedback.

[0551] "Model improvement methods" are means of improving the accuracy of a model based on user feedback.

[0552] This invention is a system that supports decision-making for busy business leaders, managers, and professionals. This system primarily consists of the following elements:

[0553] 1. Data acquisition methods

[0554] 2. Data preprocessing means

[0555] 3. Learning methods

[0556] 4. Recommended generation method

[0557] 5. User Feedback Methods

[0558] 6. Model Improvement Methods

[0559] Data acquisition methods

[0560] The server retrieves data on the user's past behavior and decision-making history from the database. For example, it collects data such as purchase history, schedules, and message exchanges. This provides the foundational data needed to understand the user's decision-making patterns.

[0561] The device monitors user actions and behaviors in real time and sends that information to the server. For example, it monitors actions such as a user adding an event to their calendar in real time and sends that data to the server.

[0562] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[0563] Data preprocessing means

[0564] The server cleanses the acquired data, removing inconsistent data. Furthermore, it normalizes the data and converts it into a format suitable for analysis. For example, it might unify date data in different formats.

[0565] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. This process ensures data consistency.

[0566] Learning methods

[0567] The server learns user decision-making patterns using deep learning models (e.g., recurrent neural networks or transformer models). Specifically, data is divided into batches and fed into the model, and learning is performed epoch by epoch.

[0568] The training process is periodically rerun in batches, and the model is continuously updated with new data. This process allows the model to make recommendations that reflect the latest data.

[0569] Recommended generation method

[0570] The server uses a pre-trained model to generate personalized recommendations based on the user's situation. For example, when a business leader is deciding which projects to allocate resources to next month, the server suggests the best options based on past success rates and cost-effectiveness.

[0571] The terminal notifies the user of recommendations from the server and displays them on the interface. Notifications are made in real time, allowing the user to check them immediately.

[0572] User Feedback Methods

[0573] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, the results are also sent to the server.

[0574] The device collects user feedback in real time and sends it to the server. The collected data is immediately reflected in the database.

[0575] Model Improvement Means

[0576] The server improves the model's accuracy using user feedback and newly collected data. By analyzing the feedback data and optimizing the model's loss function, the accuracy of the next recommendation improves.

[0577] Specific examples and prompt statements

[0578] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0579] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[0580] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[0581] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[0582] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[0583] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[0584] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0585] Examples of prompts for a generative AI model:

[0586] "Based on past project data, please recommend the optimal resource allocation for the next month. Consider variables such as success rate, cost, and time."

[0587] By using the above system, it becomes possible to support the user's decision-making process and provide personalized recommendations, thereby improving productivity and efficiency.

[0588] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0589] Step 1: Data Collection

[0590] The server retrieves the user's past behavioral data and decision-making history from the database. Specifically, it issues queries to collect data such as purchase history, schedules, and message exchanges. The input is the query results from the database, and the output is the set of collected data.

[0591] The device monitors user actions and behaviors in real time and sends that information to the server. Specifically, it captures actions such as a user adding an event to their calendar and sends that data to the server. The input is the user's real-time actions, and the output is the monitored data.

[0592] As part of the initial setup in the form of a questionnaire, users enter their values ​​and priorities. This data is also sent to the server and stored in the database. The input is the user's questionnaire responses, and the output is the data sent to the server.

[0593] Step 2: Data Preprocessing

[0594] The server cleanses the acquired data and removes inconsistent data. Specifically, it imputes missing values ​​and removes outliers. The input is the collected raw data, and the output is the cleansed data.

[0595] The server further normalizes the data and standardizes its format. For example, it unifies date data in different formats. The input is cleansed data, and the output is normalized data.

[0596] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. The input is the user's survey data, and the output is the converted data.

[0597] Step 3: Model Training

[0598] The server trains a deep learning model using normalized data. Specifically, it batch processes the data using recurrent neural networks or transformer models and performs training in epoch units. The input is a normalized dataset, and the output is the trained model.

[0599] The training process is periodically rerun in batches, and the model is continuously updated using newly collected data. The input is the new data, and the output is the updated model.

[0600] Step 4: Recommended Generation

[0601] The server uses a trained model to generate user-specific recommendations. For example, it might suggest the optimal project resource allocation for a business leader for the following month. The input is the trained model and current user data, and the output is the generated recommendations.

[0602] The terminal notifies the user of recommendations received from the server and displays them on the interface. Specifically, it presents recommendations to the user through pop-up notifications. The input is the recommendations, and the output is the notification to the user.

[0603] Step 5: User Feedback

[0604] Users review the provided recommendations and make corrections or provide feedback as needed. For example, they may be given options to accept, reject, or modify the recommendations. The input is the user's feedback, and the output is the modified or accepted data.

[0605] The device collects user feedback in real time and sends it to the server. The input is the user's real-time feedback, and the output is the data sent to the server.

[0606] Step 6: Model Improvement

[0607] The server improves the model's accuracy using user feedback and newly collected data. Specifically, it analyzes the feedback data and optimizes the model's loss function. The input is the feedback data and newly collected data, and the output is the updated model.

[0608] By following these steps, the system can continue to provide recommendations optimized for the user's situation.

[0609] (Application Example 1)

[0610] 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."

[0611] In conventional factory robot operations, it was difficult to efficiently manage the operating data, maintenance records, and production plans of individual robots, making it challenging to determine the optimal operating schedule and maintenance timing. As a result, production efficiency decreased and costs increased due to unnecessary maintenance. The present invention aims to solve these problems and provide a new system for optimizing robot operations within a factory.

[0612] 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.

[0613] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, means for acquiring and analyzing data to optimize the operational efficiency of robots in a factory, means for recommending an optimal operational schedule and maintenance timing, and means for displaying the recommended operational schedule and maintenance timing. This enables efficient operational management and appropriate maintenance of factory robots.

[0614] "Data collection means" refers to means for acquiring information such as users' past behavioral data, decision-making history, robot operation data within the factory, maintenance records, and production plans.

[0615] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, and further converting it into a format suitable for analyzing the operational efficiency of robots within a factory.

[0616] "Learning methods" refer to techniques that improve the accuracy of recommendations by using deep learning models based on user and robot behavior data.

[0617] A "recommendation generation method" is a means of generating recommendations to optimize the next task or maintenance timing of a factory robot using a pre-trained model.

[0618] A "user feedback mechanism" is a means of collecting user feedback and modifications to generated recommendations and incorporating them into the system.

[0619] "Model improvement methods" refer to techniques for continuously improving the accuracy of a learning model using user feedback and new data.

[0620] "Means of acquiring and analyzing data to optimize operational efficiency" refers to methods of collecting data on robots within a factory and analyzing it to achieve optimal operational efficiency.

[0621] "Means for recommending optimal operating schedules and maintenance timings" refers to methods for proposing optimal operating schedules and maintenance timings for robots to users based on collected and analyzed data.

[0622] "Means for displaying recommended operational schedules and maintenance timings" refers to a means of visually communicating the generated recommendations by displaying them on the user's device.

[0623] This invention is a system for optimizing the operational efficiency of robots in a factory, and its specific form is shown below.

[0624] System Configuration

[0625] The system of the present invention consists of the following elements:

[0626] 1. Data acquisition methods

[0627] 2. Data preprocessing means

[0628] 3. Learning methods

[0629] 4. Recommended generation method

[0630] 5. User Feedback Methods

[0631] 6. Model Improvement Methods

[0632] 7. Means for acquiring and analyzing data to optimize operational efficiency

[0633] 8. Means for recommending optimal operating schedules and maintenance timings.

[0634] 9. Means for displaying recommended operational schedules and maintenance timings.

[0635] System operation

[0636] Data acquisition methods

[0637] The server retrieves operational data, maintenance records, and production plans from robots within the factory via an API.

[0638] Users provide past behavioral data and decision-making history, which is also sent to the server and stored in the database.

[0639] Data preprocessing means

[0640] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analyzing operational efficiency within the factory.

[0641] The terminal converts the data obtained from the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[0642] Learning methods

[0643] The server uses deep learning models (e.g., recurrent neural networks or transformer models) to learn user and robot behavior data and incorporate it into the model.

[0644] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[0645] Recommended generation method

[0646] The server uses pre-trained models to recommend optimizations for the next tasks and maintenance timings of factory robots.

[0647] The terminal notifies the user of recommendations from the server and displays them on the interface.

[0648] User Feedback Methods

[0649] Users review the provided recommendations and make corrections or provide feedback. This feedback is sent to the server.

[0650] The device collects user feedback in real time and sends it to the server.

[0651] Model Improvement Means

[0652] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[0653] Specific example

[0654] As a concrete example, consider a scenario where a factory manager decides which robots should be assigned which tasks in the following month.

[0655] 1. Data collection method: The server acquires historical data such as the operation data, success rate, maintenance records, and production plans of factory robots.

[0656] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the data into a format for analyzing the robot's operational efficiency and sends it to the server.

[0657] 3. Learning method: The server uses a deep learning model (e.g., Keras / TensorFlow) to learn the robot's operational patterns.

[0658] 4. Recommendation generation method: The server generates recommendations for the optimal task assignment for the following month and notifies the manager via the terminal.

[0659] 5. User Feedback Method: Managers review, approve, or revise recommendations. Feedback is sent to the server via the terminal.

[0660] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0661] This optimizes the operational schedule and maintenance timing of robots within the factory, improving productivity and efficiency.

[0662] Example of a prompt

[0663] "Please retrieve robot operation data from the factory data API."

[0664] Please normalize your data using Pandas and Scikit-learn.

[0665] "Load the saved Keras model and perform inference."

[0666] "Generate recommendations based on the model's inference results."

[0667] "Please display the generated recommendations in JSON format."

[0668] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0669] Step 1:

[0670] The server retrieves information such as robot operation data, maintenance records, and production plans from robots within the factory via an API. This includes each robot's work history, current status, and future work schedule. Input is raw data retrieved from the API, and output is canonical data stored in the database.

[0671] Step 2:

[0672] The server cleanses the acquired raw data and removes inconsistent data. Next, it normalizes the data and converts it into a format suitable for analyzing the operational efficiency of factory robots. Data processing libraries such as Pandas and Scikit-learn are used for this. The input is the raw data acquired in step 1, and the output is the cleansed and normalized data.

[0673] Step 3:

[0674] The server uses a deep learning model (e.g., Keras / TensorFlow) to train on normalized data. The training process involves pattern extraction and model training using recurrent neural networks or transformer models based on acquired historical data. The input is the data preprocessed in step 2, and the output is the trained model.

[0675] Step 4:

[0676] The server uses a pre-trained model to perform inference and recommend optimizations for the next task and maintenance timing of factory robots. This involves converting the generated recommendations into JSON format and sending them to the terminal. The input is the pre-trained model and current data, and the output is the recommendations in JSON format.

[0677] Step 5:

[0678] The terminal notifies the user of recommendations received from the server and displays them on the interface. This allows the user to review the recommendations and provide feedback or make corrections as needed. The input is the recommendations received from the server in JSON format, and the output is the recommendations displayed in a format usable by the user.

[0679] Step 6:

[0680] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The feedback is sent to the server via the device. The input is the recommendations displayed on the device, and the output is the feedback sent to the server.

[0681] Step 7:

[0682] The server uses received user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided. The input is user feedback and new data, and the output is the updated trained model.

[0683] 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.

[0684] This invention is a system that supports decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes the user's emotions. Its specific form is described below.

[0685] System Configuration

[0686] The system of the present invention consists of the following elements:

[0687] 1. Data acquisition methods

[0688] 2. Data preprocessing means

[0689] 3. Learning methods

[0690] 4. Recommended generation method

[0691] 5. User Feedback Methods

[0692] 6. Model Improvement Methods

[0693] 7. Emotional Engine

[0694] System operation

[0695] Data acquisition methods

[0696] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[0697] The device monitors the user's real-time operations and actions and sends that data to the server.

[0698] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[0699] The emotion engine acquires emotion data from the user's voice, facial expressions, text, etc., and sends it to the server.

[0700] Data preprocessing means

[0701] The server cleanses all acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[0702] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[0703] The emotion data acquired by the emotion engine is also appropriately preprocessed before being sent to the server.

[0704] Learning methods

[0705] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process.

[0706] The trained model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[0707] Recommended generation method

[0708] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it might recommend which project to allocate resources to next, taking into account whether the user is feeling uplifted or stressed.

[0709] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[0710] User Feedback Methods

[0711] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide feedback regarding their own feelings.

[0712] The device collects user feedback in real time and sends that data to the server.

[0713] Model Improvement Means

[0714] The server uses user feedback, new behavioral data, and sentiment data to retrain the model, aiming for personalization and improved accuracy.

[0715] This will improve the quality of future recommendations and more effectively support user decision-making.

[0716] Specific example

[0717] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0718] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data. The emotion engine collects emotional data from the leader's voice and facial expressions.

[0719] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[0720] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[0721] 4. Recommended generation method: The server recommends high-risk projects when the user is feeling exhilarated, and stable projects when the user is feeling stressed.

[0722] 5. User Feedback Methods: Leaders review, approve, or revise recommendations. They also provide emotional feedback.

[0723] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0724] This enables the system of the present invention to support the user's decision-making process and provide personalized recommendations that take emotions into consideration. This, in turn, improves user productivity and efficiency and contributes to stress reduction.

[0725] The following describes the processing flow.

[0726] ---

[0727] Step 1:

[0728] The server retrieves the user's past behavioral data and decision-making history from the database. This includes purchase history, project evaluations, and the results of past decisions.

[0729] The device monitors the user's real-time actions and behavior and sends that data to the server. For example, it might track which web pages the user is browsing or which apps they are using.

[0730] The emotion engine acquires emotional data from the user's voice, facial expressions, and text. For example, it analyzes the user's stress level and happiness level from the tone of their speech and facial expressions.

[0731] Step 2:

[0732] The server cleanses all acquired data, removing inaccurate data and noise. For example, it removes duplicate and inconsistent data.

[0733] The server performs normalization and transforms the data into a format suitable for analysis. For example, it scales numerical data or encodes categorical data.

[0734] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server. For example, it extracts text data from the survey form and converts it into JSON format.

[0735] The emotion engine preprocesses emotion data and sends it to the server. For example, it converts audio data to text and assigns emotion labels to it.

[0736] Step 3:

[0737] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process.

[0738] The training process is periodically rerun in batches, and the model is continuously updated with new data. For example, the model is retrained every night with new data.

[0739] Step 4:

[0740] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is feeling stressed, it will recommend safer options.

[0741] The device notifies the user of recommendations received from the server and displays them on the interface. For example, it might use smartphone push notifications or notifications from wearable devices.

[0742] Step 5:

[0743] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[0744] The device collects user feedback in real time and sends that data to the server. For example, when a user presses the "Approve" button in a push notification, that selection is recorded and sent.

[0745] Step 6:

[0746] The server uses user feedback and new behavioral and sentiment data to retrain its model, aiming for personalization and improved accuracy.

[0747] This improves the quality of future recommendations and more effectively supports user decision-making. For example, new recommendations can be readjusted based on the user's previous acceptance of a recommendation and their subsequent emotional response.

[0748] ---

[0749] The above describes the specific processing flow in the system of the present invention that incorporates an emotion engine. This enables the user's decision-making process to be supported efficiently and consistently while taking emotions into consideration.

[0750] (Example 2)

[0751] 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".

[0752] Busy business leaders, managers, and professionals face a challenge in making quick and accurate decisions, as they are often influenced by emotions. This can lead to stress and decreased productivity. Therefore, there is a need for systems that support decision-making by providing personalized recommendations while considering the user's emotional state.

[0753] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0754] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and sentiment data acquisition means. This makes it possible to provide personalized and sentiment-responsive recommendations based on user behavior data and sentiment data.

[0755] "Data collection means" refers to means for acquiring users' past behavioral data and decision-making history and transmitting it to a server.

[0756] "Data preprocessing means" refers to means for cleansing and normalizing acquired data and converting it into a format suitable for analysis.

[0757] "Learning methods" refer to techniques for performing deep learning using recurrent neural networks (RNNs) or transformer models based on user behavior data, values, and emotional data.

[0758] A "recommendation generation method" is a means of generating recommendations based on the current user's situation and emotions, using a pre-trained model.

[0759] A "user feedback mechanism" is a means for users to provide feedback on the recommendations they have been given, to collect that feedback in real time, and to send it to the server.

[0760] "Model improvement methods" refer to techniques for retraining a model using user feedback, new behavioral data, and sentiment data, with the aim of improving personalization and accuracy.

[0761] "Means for acquiring emotional data" refers to a means of collecting emotional data from the user's voice, facial expressions, and text, and transmitting it to a server.

[0762] This invention provides a system to support decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes user emotions. The system of this invention includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and emotion data acquisition means. Its specific form is shown below.

[0763] 1. Data acquisition methods

[0764] The server retrieves the user's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. For example, the server uses an API to collect this data.

[0765] The device monitors the user's real-time actions and behaviors, including the location and frequency of clicks and taps, and sends this data to a server. This is done using the device's sensors.

[0766] As part of the initial setup, users enter their values ​​and priorities into the system via web forms or in-app questionnaires. This input data is sent to the server via the device and stored in a database.

[0767] The emotion engine acquires emotional data from the user's voice, facial expressions, text, etc., and sends it to the server. For example, it performs facial recognition through the camera and voice analysis through the microphone.

[0768] 2. Data preprocessing means

[0769] The server cleanses all acquired data, removing inaccuracies and noise. Specifically, it removes duplicate data and imputes missing values. It also normalizes the data, converting data from different data sources into a unified format.

[0770] The terminal converts the survey results entered by the user into an appropriate data format, such as JSON or XML, and sends that data to the server.

[0771] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[0772] 3. Learning methods

[0773] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The model is trained on a training dataset, and its parameters are adjusted. The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[0774] 4. Recommended generation method

[0775] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[0776] The device displays recommendations received from the server to the user via an interface. Information is provided in real time using notifications and pop-up messages.

[0777] 5. User Feedback Methods

[0778] Users can review the provided recommendations and approve, reject, or modify them. They can also provide feedback on whether the recommendations were appropriate and whether they fit their emotional state.

[0779] The device collects user feedback in real time and sends the results to the server.

[0780] 6. Model Improvement Methods

[0781] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[0782] Specific example

[0783] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0784] The server retrieves historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[0785] The emotion engine collects emotional data from the leader's voice and facial expressions.

[0786] The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[0787] The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[0788] The server recommends high-risk projects when you're feeling exhilarated, and stable projects when you're feeling stressed.

[0789] Leaders review, approve, or revise recommendations, and also provide emotional feedback.

[0790] The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0791] This enables the system of the present invention to support user decision-making and provide personalized recommendations that take emotions into consideration. Furthermore, this improves user productivity and efficiency and contributes to stress reduction.

[0792] Example of a prompt

[0793] The system generates an answer to the question, "Which project should we allocate resources to next month?" by considering the success rate of recent projects, resource utilization, and leader emotional data (emotional state, stress level, etc.).

[0794] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0795] Step 1:

[0796] Data collection:

[0797] The server retrieves users' past behavioral data and decision-making history from the database using multiple APIs. This includes purchase history, schedules, and message exchanges.

[0798] Input: User ID

[0799] Output: User data such as purchase history, schedule, and message history.

[0800] The device monitors the user's real-time actions (location and frequency of clicks and taps) and behavior, and periodically sends this data to the server. Activity data is captured using the device's sensors.

[0801] Input: User operation data

[0802] Output: Real-time operation data

[0803] As part of the initial setup, users enter their values ​​and priorities through web forms or in-app questionnaires. This data is transmitted to the server via the device and stored in a database.

[0804] Input: Survey data

[0805] Output: Values ​​and priority data

[0806] The emotion engine acquires emotional data from the user's voice, facial expressions, and text, and sends it to the server. For example, it recognizes facial expressions through the camera and performs voice analysis through the microphone.

[0807] Input: Voice, facial expressions, text data

[0808] Output: Sentiment data

[0809] Step 2:

[0810] Data preprocessing:

[0811] The server cleanses the acquired data, removing inaccuracies and noise. Specific tasks include deleting duplicate data and imputing missing values. It also normalizes the data, converting data from different sources into a unified format.

[0812] Input: Raw data (purchase history, schedule, message history, real-time operation data, sentiment data)

[0813] Output: Cleansed and normalized data

[0814] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[0815] Input: Survey results

[0816] Output: Survey data in JSON or XML format

[0817] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[0818] Input: Sentiment data

[0819] Output: Preprocessed sentiment data

[0820] Step 3:

[0821] Model learning:

[0822] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process. The model is trained on a training dataset and its parameters are adjusted.

[0823] Input: Cleansed and normalized data

[0824] Output: Trained model

[0825] The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[0826] Input: Training data

[0827] Output: High-precision model

[0828] Step 4:

[0829] Recommended generation:

[0830] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[0831] Input: Trained model, current situation data, sentiment data

[0832] Output: Recommendations

[0833] The terminal notifies the user of recommendations received from the server and displays them on the interface. Information is provided in real time using notifications and pop-up messages.

[0834] Input: Recommendations

[0835] Output: Display on the user screen

[0836] Step 5:

[0837] Collecting user feedback:

[0838] Users review the provided recommendations and can approve, reject, or modify them. They also provide feedback on whether the recommendations are appropriate and whether they align with their emotional state.

[0839] Input: User feedback on recommendations

[0840] Output: Revised recommendations, feedback data

[0841] The device collects user feedback in real time and sends the results to the server.

[0842] Input: User feedback data

[0843] Output: Feedback data sent to the server

[0844] Step 6:

[0845] Model improvements:

[0846] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[0847] Input: Feedback data, new behavioral data, emotional data

[0848] Output: Improved model

[0849] This allows the system to support user decision-making, provide personalized recommendations that take emotions into account, and improve user productivity and efficiency, as well as reduce stress.

[0850] (Application Example 2)

[0851] 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."

[0852] In modern factories and production sites, decision-making regarding robot operation and maintenance is crucial. However, many systems fail to consider the emotions and stress levels of managers, leading to inefficient resource allocation and increased robot failure risks. Therefore, decision support systems that take managers' emotions into account need to provide more optimal resource allocation and maintenance schedules.

[0853] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means. This makes it possible to grasp the emotional state of the administrator in real time and recommend appropriate resource allocation and maintenance schedules based on that.

[0854] "Data collection means" refers to methods for acquiring users' past behavioral data and decision-making history, as well as real-time operation data and emotional data.

[0855] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, removing inaccurate data and noise, and converting it into a format suitable for analysis.

[0856] "Learning methods" refer to techniques that use deep learning models or machine learning algorithms to model user decision-making patterns based on user behavior data, decision history, values ​​data, sentiment data, etc.

[0857] A "recommendation generation method" is a means of generating appropriate recommendations based on the current user situation and sentiment data, using a trained model.

[0858] A "user feedback mechanism" refers to a means of collecting feedback from users, such as modifications or approvals of recommendations provided by users.

[0859] "Model improvement methods" refer to techniques for retraining a model using user feedback and new behavioral and sentiment data to improve the accuracy of recommendations.

[0860] "Emotion recognition means" refers to methods for acquiring and analyzing emotional data from a user's voice, facial expressions, text, etc.

[0861] A "real-time decision support system" is a means of providing appropriate recommendations immediately based on data and sentiment data collected in real time, thereby supporting the user's decision-making.

[0862] The embodiments for carrying out this invention will be described below. The factory robot management assistant system includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means.

[0863] System Configuration

[0864] Data acquisition methods

[0865] In this system, the server retrieves the administrator's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. The terminal monitors real-time operation data and sends it to the server. The emotion engine, used as an emotion recognition tool, retrieves emotion data from the administrator's voice, facial expressions, text, etc., and sends it to the server.

[0866] Data preprocessing means

[0867] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis. On the terminal side, the survey results entered by the user are converted into an appropriate data format (e.g., JSON or XML) and sent to the server. Sentiment data is also preprocessed and sent to the server.

[0868] Learning methods

[0869] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The learned model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[0870] Recommended generation method

[0871] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it recommends high-risk projects when the administrator is emotionally uplifted, and stable projects when they are stressed. The terminal notifies the user of the recommendations received from the server and displays them in the interface.

[0872] User Feedback Methods

[0873] Users review the provided recommendations and provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide emotional feedback. The device collects user feedback in real time and sends the data to the server.

[0874] Model Improvement Means

[0875] The server retrains its model using user feedback and new behavioral and sentiment data to improve the accuracy of its recommendations.

[0876] Hardware and software to be used

[0877] Hardware: Computer, camera, and microphone built into the factory robot.

[0878] Software: Python, scikit-learn, emotion_recognition library

[0879] Processing Overview

[0880] 1. Data collection:

[0881] The server retrieves the administrator's past behavioral data and decision-making history from a database, while the terminal monitors real-time operation data. The emotion engine collects emotional data from the administrator's voice and facial expressions.

[0882] 2. Data preprocessing:

[0883] The acquired data is cleansed and normalized by the server. Data such as survey results are also converted into an appropriate format.

[0884] 3. Learning:

[0885] The server uses a deep learning model to learn from the administrator's behavioral and emotional data.

[0886] 4. Recommended generation:

[0887] Based on real-time data and pre-trained models, it generates recommendations tailored to the current administrator situation.

[0888] 5. Gathering feedback:

[0889] User feedback is sent to the server via the device.

[0890] 6. Model Improvement:

[0891] The server retrains the model based on the feedback, improving the accuracy of future recommendations.

[0892] Specific example

[0893] For example, if a factory robot manager is under a very high level of stress, this management assistant app will recommend reducing the high-risk tasks performed by the robots. Conversely, if the manager is emotionally stable, it will prioritize efficiency and make recommendations that include high-risk tasks.

[0894] Example of a prompt

[0895] "Based on the following data, please generate the optimal work schedule for factory robot management. The data is as follows:"

[0896] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0897] Step 1:

[0898] The server retrieves the administrator's past behavioral data and decision history from the database. This includes information such as purchase history, schedules, and message exchanges. The input is past behavioral data and decision history from the database, and the output is a dataset containing this data. Specifically, it uses SQL queries to retrieve the necessary information from the database.

[0899] Step 2:

[0900] The terminal monitors real-time operation and emotion data and sends it to the server. Inputs are administrator operations and voice and facial expression data collected by the emotion engine, while output is real-time data sent to the server. Specifically, emotion recognition software is used to extract emotion information from voice and facial expressions.

[0901] Step 3:

[0902] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data. The input is the data obtained in steps 1 and 2, and the output is the cleansed and normalized dataset. Specifically, it uses a data cleaning tool to remove incomplete data and noise, and a standard scaler to normalize the data.

[0903] Step 4:

[0904] The server performs deep learning using recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and sentiment data. The input is a cleansed and normalized dataset, and the output is the trained recommendation model. Specifically, the model is trained using deep learning frameworks such as PyTorch and TensorFlow.

[0905] Step 5:

[0906] The server uses a pre-trained model to generate recommendations based on the current user's situation and sentiment. The input is real-time interaction and sentiment data, and the output is recommendations. Specifically, it inputs real-time data into the pre-trained model and generates recommendations.

[0907] Step 6:

[0908] The terminal notifies the user of recommendations received from the server and displays them on the interface. The input is the recommendations from the server, and the output is the user's feedback. Specifically, it displays the recommendations using the user interface (UI) and receives feedback from the user.

[0909] Step 7:

[0910] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The input is the recommendations displayed on the device, and the output is the feedback sent to the device. Specifically, the user approves, rejects, or modifies the recommendations and inputs the result into the device.

[0911] Step 8:

[0912] The server retrains the model using user feedback and new behavioral and sentiment data. The input is user feedback data and new data, and the output is the updated trained model. Specifically, it retrains the model using feedback data to improve accuracy.

[0913] 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.

[0914] 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.

[0915] 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.

[0916] [Third Embodiment]

[0917] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0918] 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.

[0919] 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).

[0920] 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.

[0921] 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.

[0922] 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).

[0923] 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.

[0924] 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.

[0925] 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.

[0926] 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.

[0927] 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.

[0928] 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".

[0929] This invention is a system that supports decision-making for busy business leaders, managers, and experts, and its specific form is shown below.

[0930] System Configuration

[0931] The system of this invention mainly consists of the following elements:

[0932] 1. Data acquisition methods

[0933] 2. Data preprocessing means

[0934] 3. Learning methods

[0935] 4. Recommended generation method

[0936] 5. User Feedback Methods

[0937] 6. Model Improvement Methods

[0938] System operation

[0939] Data acquisition methods

[0940] The server retrieves the user's past behavioral data and decision-making history from a database. This includes purchase history, schedules, and message exchanges.

[0941] The terminal monitors user operations and actions in real time and sends this information to the server.

[0942] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This information is also sent to the server and stored in the database.

[0943] Data preprocessing means

[0944] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analysis.

[0945] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[0946] Learning methods

[0947] The server learns the user's decision-making process using deep learning models (e.g., recurrent neural networks or transformer models). This allows it to extract individual user patterns and incorporate them into the model.

[0948] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[0949] Recommended generation method

[0950] The server uses a pre-trained model to generate user-specific recommendations. For example, when a user is deciding on project resource allocation, it suggests the best option based on past success rates and cost-effectiveness.

[0951] The terminal notifies the user of recommendations from the server and displays them on the interface.

[0952] User Feedback Methods

[0953] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, they also send the results to the server.

[0954] The device collects user feedback in real time and sends it to the server.

[0955] Model Improvement Means

[0956] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[0957] Specific example

[0958] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[0959] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[0960] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[0961] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[0962] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[0963] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[0964] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[0965] In this way, the system of the present invention supports the user's decision-making process and provides optimized decision-making. This enables personalized support based on individual values, improving the user's productivity and efficiency.

[0966] The following describes the processing flow.

[0967] ---

[0968] Step 1:

[0969] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[0970] The device monitors the user's real-time operations and actions and sends that data to the server.

[0971] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server.

[0972] Step 2:

[0973] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[0974] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[0975] Step 3:

[0976] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data and values. This allows it to model the user's decision-making process.

[0977] The trained model reflects the user's characteristics and past decision-making patterns.

[0978] Step 4:

[0979] The server uses a pre-trained model to generate recommendations based on the user's current situation. For example, it might recommend which project the user should allocate resources to next.

[0980] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[0981] Step 5:

[0982] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[0983] The device collects user feedback in real time and sends that data to the server.

[0984] Step 6:

[0985] The server uses user feedback and new behavioral data to retrain the model, aiming for personalization and improved accuracy.

[0986] This will improve the quality of future recommendations and more effectively support user decision-making.

[0987] ---

[0988] The above describes the specific processing flow in the system of the present invention. This enables efficient and consistent support for the user's decision-making process.

[0989] (Example 1)

[0990] 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."

[0991] Traditional decision support systems have faced challenges in effectively assisting busy business leaders and managers with their decision-making. In particular, their inability to provide personalized recommendations that reflect individual user values ​​and priorities often leads to a decline in recommendation quality and reduced decision-making efficiency. Furthermore, the difficulty in effectively incorporating user feedback to improve model accuracy results in a gradual deterioration of recommendation quality.

[0992] 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.

[0993] In this invention, the server includes data collection means for acquiring the user's past behavioral data and decision-making history, data preprocessing means for cleansing and normalizing the acquired data, learning means using a deep learning model, recommendation generation means for generating recommendations, user feedback means for receiving user feedback, and model improvement means for improving the accuracy of the model. This makes it possible to provide personalized recommendations based on the user's individual behavior and values, and to continuously improve the quality of recommendations and the efficiency of decision-making.

[0994] "Data collection methods" refer to means of acquiring users' past behavioral data and decision-making history.

[0995] "Data preprocessing means" refers to methods for cleansing and normalizing acquired data.

[0996] A "learning method" is a means of learning user decision-making patterns using a deep learning model.

[0997] A "recommendation generation method" is a method for generating recommendations for users using a pre-trained model.

[0998] A "user feedback method" is a means of receiving user feedback.

[0999] "Model improvement methods" are means of improving the accuracy of a model based on user feedback.

[1000] This invention is a system that supports decision-making for busy business leaders, managers, and professionals. This system primarily consists of the following elements:

[1001] 1. Data acquisition methods

[1002] 2. Data preprocessing means

[1003] 3. Learning methods

[1004] 4. Recommended generation method

[1005] 5. User Feedback Methods

[1006] 6. Model Improvement Methods

[1007] Data acquisition methods

[1008] The server retrieves data on the user's past behavior and decision-making history from the database. For example, it collects data such as purchase history, schedules, and message exchanges. This provides the foundational data needed to understand the user's decision-making patterns.

[1009] The device monitors user actions and behaviors in real time and sends that information to the server. For example, it monitors actions such as a user adding an event to their calendar in real time and sends that data to the server.

[1010] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[1011] Data preprocessing means

[1012] The server cleanses the acquired data, removing inconsistent data. Furthermore, it normalizes the data and converts it into a format suitable for analysis. For example, it might unify date data in different formats.

[1013] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. This process ensures data consistency.

[1014] Learning methods

[1015] The server learns user decision-making patterns using deep learning models (e.g., recurrent neural networks or transformer models). Specifically, data is divided into batches and fed into the model, and learning is performed epoch by epoch.

[1016] The training process is periodically rerun in batches, and the model is continuously updated with new data. This process allows the model to make recommendations that reflect the latest data.

[1017] Recommended generation method

[1018] The server uses a pre-trained model to generate personalized recommendations based on the user's situation. For example, when a business leader is deciding which projects to allocate resources to next month, the server suggests the best options based on past success rates and cost-effectiveness.

[1019] The terminal notifies the user of recommendations from the server and displays them on the interface. Notifications are made in real time, allowing the user to check them immediately.

[1020] User Feedback Methods

[1021] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, the results are also sent to the server.

[1022] The device collects user feedback in real time and sends it to the server. The collected data is immediately reflected in the database.

[1023] Model Improvement Means

[1024] The server improves the model's accuracy using user feedback and newly collected data. By analyzing the feedback data and optimizing the model's loss function, the accuracy of the next recommendation improves.

[1025] Specific examples and prompt statements

[1026] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[1027] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[1028] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[1029] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[1030] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[1031] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[1032] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1033] Examples of prompts for a generative AI model:

[1034] "Based on past project data, please recommend the optimal resource allocation for the next month. Consider variables such as success rate, cost, and time."

[1035] By using the above system, it becomes possible to support the user's decision-making process and provide personalized recommendations, thereby improving productivity and efficiency.

[1036] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1037] Step 1: Data Collection

[1038] The server retrieves the user's past behavioral data and decision-making history from the database. Specifically, it issues queries to collect data such as purchase history, schedules, and message exchanges. The input is the query results from the database, and the output is the set of collected data.

[1039] The device monitors user actions and behaviors in real time and sends that information to the server. Specifically, it captures actions such as a user adding an event to their calendar and sends that data to the server. The input is the user's real-time actions, and the output is the monitored data.

[1040] As part of the initial setup in the form of a questionnaire, users enter their values ​​and priorities. This data is also sent to the server and stored in the database. The input is the user's questionnaire responses, and the output is the data sent to the server.

[1041] Step 2: Data Preprocessing

[1042] The server cleanses the acquired data and removes inconsistent data. Specifically, it imputes missing values ​​and removes outliers. The input is the collected raw data, and the output is the cleansed data.

[1043] The server further normalizes the data and standardizes its format. For example, it unifies date data in different formats. The input is cleansed data, and the output is normalized data.

[1044] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. The input is the user's survey data, and the output is the converted data.

[1045] Step 3: Model Training

[1046] The server trains a deep learning model using normalized data. Specifically, it batch processes the data using recurrent neural networks or transformer models and performs training in epoch units. The input is a normalized dataset, and the output is the trained model.

[1047] The training process is periodically rerun in batches, and the model is continuously updated using newly collected data. The input is the new data, and the output is the updated model.

[1048] Step 4: Recommended Generation

[1049] The server uses a trained model to generate user-specific recommendations. For example, it might suggest the optimal project resource allocation for a business leader for the following month. The input is the trained model and current user data, and the output is the generated recommendations.

[1050] The terminal notifies the user of recommendations received from the server and displays them on the interface. Specifically, it presents recommendations to the user through pop-up notifications. The input is the recommendations, and the output is the notification to the user.

[1051] Step 5: User Feedback

[1052] Users review the provided recommendations and make corrections or provide feedback as needed. For example, they may be given options to accept, reject, or modify the recommendations. The input is the user's feedback, and the output is the modified or accepted data.

[1053] The device collects user feedback in real time and sends it to the server. The input is the user's real-time feedback, and the output is the data sent to the server.

[1054] Step 6: Model Improvement

[1055] The server improves the model's accuracy using user feedback and newly collected data. Specifically, it analyzes the feedback data and optimizes the model's loss function. The input is the feedback data and newly collected data, and the output is the updated model.

[1056] By following these steps, the system can continue to provide recommendations optimized for the user's situation.

[1057] (Application Example 1)

[1058] 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."

[1059] In conventional factory robot operations, it was difficult to efficiently manage the operating data, maintenance records, and production plans of individual robots, making it challenging to determine the optimal operating schedule and maintenance timing. As a result, production efficiency decreased and costs increased due to unnecessary maintenance. The present invention aims to solve these problems and provide a new system for optimizing robot operations within a factory.

[1060] 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.

[1061] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, means for acquiring and analyzing data to optimize the operational efficiency of robots in a factory, means for recommending an optimal operational schedule and maintenance timing, and means for displaying the recommended operational schedule and maintenance timing. This enables efficient operational management and appropriate maintenance of factory robots.

[1062] "Data collection means" refers to means for acquiring information such as users' past behavioral data, decision-making history, robot operation data within the factory, maintenance records, and production plans.

[1063] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, and further converting it into a format suitable for analyzing the operational efficiency of robots within a factory.

[1064] "Learning methods" refer to techniques that improve the accuracy of recommendations by using deep learning models based on user and robot behavior data.

[1065] A "recommendation generation method" is a means of generating recommendations to optimize the next task or maintenance timing of a factory robot using a pre-trained model.

[1066] A "user feedback mechanism" is a means of collecting user feedback and modifications to generated recommendations and incorporating them into the system.

[1067] "Model improvement methods" refer to techniques for continuously improving the accuracy of a learning model using user feedback and new data.

[1068] "Means of acquiring and analyzing data to optimize operational efficiency" refers to methods of collecting data on robots within a factory and analyzing it to achieve optimal operational efficiency.

[1069] "Means for recommending optimal operating schedules and maintenance timings" refers to methods for proposing optimal operating schedules and maintenance timings for robots to users based on collected and analyzed data.

[1070] "Means for displaying recommended operational schedules and maintenance timings" refers to a means of visually communicating the generated recommendations by displaying them on the user's device.

[1071] This invention is a system for optimizing the operational efficiency of robots in a factory, and its specific form is shown below.

[1072] System Configuration

[1073] The system of the present invention consists of the following elements:

[1074] 1. Data acquisition methods

[1075] 2. Data preprocessing means

[1076] 3. Learning methods

[1077] 4. Recommended generation method

[1078] 5. User Feedback Methods

[1079] 6. Model Improvement Methods

[1080] 7. Means for acquiring and analyzing data to optimize operational efficiency

[1081] 8. Means for recommending optimal operating schedules and maintenance timings.

[1082] 9. Means for displaying recommended operational schedules and maintenance timings.

[1083] System operation

[1084] Data acquisition methods

[1085] The server retrieves operational data, maintenance records, and production plans from robots within the factory via an API.

[1086] Users provide past behavioral data and decision-making history, which is also sent to the server and stored in the database.

[1087] Data preprocessing means

[1088] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analyzing operational efficiency within the factory.

[1089] The terminal converts the data obtained from the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[1090] Learning methods

[1091] The server uses deep learning models (e.g., recurrent neural networks or transformer models) to learn user and robot behavior data and incorporate it into the model.

[1092] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[1093] Recommended generation method

[1094] The server uses pre-trained models to recommend optimizations for the next tasks and maintenance timings of factory robots.

[1095] The terminal notifies the user of recommendations from the server and displays them on the interface.

[1096] User Feedback Methods

[1097] Users review the provided recommendations and make corrections or provide feedback. This feedback is sent to the server.

[1098] The device collects user feedback in real time and sends it to the server.

[1099] Model Improvement Means

[1100] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[1101] Specific example

[1102] As a concrete example, consider a scenario where a factory manager decides which robots should be assigned which tasks in the following month.

[1103] 1. Data collection method: The server acquires historical data such as the operation data, success rate, maintenance records, and production plans of factory robots.

[1104] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the data into a format for analyzing the robot's operational efficiency and sends it to the server.

[1105] 3. Learning method: The server uses a deep learning model (e.g., Keras / TensorFlow) to learn the robot's operational patterns.

[1106] 4. Recommendation generation method: The server generates recommendations for the optimal task assignment for the following month and notifies the manager via the terminal.

[1107] 5. User Feedback Method: Managers review, approve, or revise recommendations. Feedback is sent to the server via the terminal.

[1108] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1109] This optimizes the operational schedule and maintenance timing of robots within the factory, improving productivity and efficiency.

[1110] Example of a prompt

[1111] "Please retrieve robot operation data from the factory data API."

[1112] Please normalize your data using Pandas and Scikit-learn.

[1113] "Load the saved Keras model and perform inference."

[1114] "Generate recommendations based on the model's inference results."

[1115] "Please display the generated recommendations in JSON format."

[1116] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1117] Step 1:

[1118] The server retrieves information such as robot operation data, maintenance records, and production plans from robots within the factory via an API. This includes each robot's work history, current status, and future work schedule. Input is raw data retrieved from the API, and output is canonical data stored in the database.

[1119] Step 2:

[1120] The server cleanses the acquired raw data and removes inconsistent data. Next, it normalizes the data and converts it into a format suitable for analyzing the operational efficiency of factory robots. Data processing libraries such as Pandas and Scikit-learn are used for this. The input is the raw data acquired in step 1, and the output is the cleansed and normalized data.

[1121] Step 3:

[1122] The server uses a deep learning model (e.g., Keras / TensorFlow) to train on normalized data. The training process involves pattern extraction and model training using recurrent neural networks or transformer models based on acquired historical data. The input is the data preprocessed in step 2, and the output is the trained model.

[1123] Step 4:

[1124] The server uses a pre-trained model to perform inference and recommend optimizations for the next task and maintenance timing of factory robots. This involves converting the generated recommendations into JSON format and sending them to the terminal. The input is the pre-trained model and current data, and the output is the recommendations in JSON format.

[1125] Step 5:

[1126] The terminal notifies the user of recommendations received from the server and displays them on the interface. This allows the user to review the recommendations and provide feedback or make corrections as needed. The input is the recommendations received from the server in JSON format, and the output is the recommendations displayed in a format usable by the user.

[1127] Step 6:

[1128] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The feedback is sent to the server via the device. The input is the recommendations displayed on the device, and the output is the feedback sent to the server.

[1129] Step 7:

[1130] The server uses received user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided. The input is user feedback and new data, and the output is the updated trained model.

[1131] 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.

[1132] This invention is a system that supports decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes the user's emotions. Its specific form is described below.

[1133] System Configuration

[1134] The system of the present invention consists of the following elements:

[1135] 1. Data acquisition methods

[1136] 2. Data preprocessing means

[1137] 3. Learning methods

[1138] 4. Recommended generation method

[1139] 5. User Feedback Methods

[1140] 6. Model Improvement Methods

[1141] 7. Emotional Engine

[1142] System operation

[1143] Data acquisition methods

[1144] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[1145] The device monitors the user's real-time operations and actions and sends that data to the server.

[1146] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[1147] The emotion engine acquires emotion data from the user's voice, facial expressions, text, etc., and sends it to the server.

[1148] Data preprocessing means

[1149] The server cleanses all acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[1150] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[1151] The emotion data acquired by the emotion engine is also appropriately preprocessed before being sent to the server.

[1152] Learning methods

[1153] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process.

[1154] The trained model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[1155] Recommended generation method

[1156] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it might recommend which project to allocate resources to next, taking into account whether the user is feeling uplifted or stressed.

[1157] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[1158] User Feedback Methods

[1159] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide feedback regarding their own feelings.

[1160] The device collects user feedback in real time and sends that data to the server.

[1161] Model Improvement Means

[1162] The server uses user feedback, new behavioral data, and sentiment data to retrain the model, aiming for personalization and improved accuracy.

[1163] This will improve the quality of future recommendations and more effectively support user decision-making.

[1164] Specific example

[1165] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[1166] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data. The emotion engine collects emotional data from the leader's voice and facial expressions.

[1167] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[1168] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[1169] 4. Recommended generation method: The server recommends high-risk projects when the user is feeling exhilarated, and stable projects when the user is feeling stressed.

[1170] 5. User Feedback Methods: Leaders review, approve, or revise recommendations. They also provide emotional feedback.

[1171] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1172] This enables the system of the present invention to support the user's decision-making process and provide personalized recommendations that take emotions into consideration. This, in turn, improves user productivity and efficiency and contributes to stress reduction.

[1173] The following describes the processing flow.

[1174] ---

[1175] Step 1:

[1176] The server retrieves the user's past behavioral data and decision-making history from the database. This includes purchase history, project evaluations, and the results of past decisions.

[1177] The device monitors the user's real-time actions and behavior and sends that data to the server. For example, it might track which web pages the user is browsing or which apps they are using.

[1178] The emotion engine acquires emotional data from the user's voice, facial expressions, and text. For example, it analyzes the user's stress level and happiness level from the tone of their speech and facial expressions.

[1179] Step 2:

[1180] The server cleanses all acquired data, removing inaccurate data and noise. For example, it removes duplicate and inconsistent data.

[1181] The server performs normalization and transforms the data into a format suitable for analysis. For example, it scales numerical data or encodes categorical data.

[1182] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server. For example, it extracts text data from the survey form and converts it into JSON format.

[1183] The emotion engine preprocesses emotion data and sends it to the server. For example, it converts audio data to text and assigns emotion labels to it.

[1184] Step 3:

[1185] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process.

[1186] The training process is periodically rerun in batches, and the model is continuously updated with new data. For example, the model is retrained every night with new data.

[1187] Step 4:

[1188] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is feeling stressed, it will recommend safer options.

[1189] The device notifies the user of recommendations received from the server and displays them on the interface. For example, it might use smartphone push notifications or notifications from wearable devices.

[1190] Step 5:

[1191] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[1192] The device collects user feedback in real time and sends that data to the server. For example, when a user presses the "Approve" button in a push notification, that selection is recorded and sent.

[1193] Step 6:

[1194] The server uses user feedback and new behavioral and sentiment data to retrain its model, aiming for personalization and improved accuracy.

[1195] This improves the quality of future recommendations and more effectively supports user decision-making. For example, new recommendations can be readjusted based on the user's previous acceptance of a recommendation and their subsequent emotional response.

[1196] ---

[1197] The above describes the specific processing flow in the system of the present invention that incorporates an emotion engine. This enables the user's decision-making process to be supported efficiently and consistently while taking emotions into consideration.

[1198] (Example 2)

[1199] 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."

[1200] Busy business leaders, managers, and professionals face a challenge in making quick and accurate decisions, as they are often influenced by emotions. This can lead to stress and decreased productivity. Therefore, there is a need for systems that support decision-making by providing personalized recommendations while considering the user's emotional state.

[1201] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1202] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and sentiment data acquisition means. This makes it possible to provide personalized and sentiment-responsive recommendations based on user behavior data and sentiment data.

[1203] "Data collection means" refers to means for acquiring users' past behavioral data and decision-making history and transmitting it to a server.

[1204] "Data preprocessing means" refers to means for cleansing and normalizing acquired data and converting it into a format suitable for analysis.

[1205] "Learning methods" refer to techniques for performing deep learning using recurrent neural networks (RNNs) or transformer models based on user behavior data, values, and emotional data.

[1206] A "recommendation generation method" is a means of generating recommendations based on the current user's situation and emotions, using a pre-trained model.

[1207] A "user feedback mechanism" is a means for users to provide feedback on the recommendations they have been given, to collect that feedback in real time, and to send it to the server.

[1208] "Model improvement methods" refer to techniques for retraining a model using user feedback, new behavioral data, and sentiment data, with the aim of improving personalization and accuracy.

[1209] "Means for acquiring emotional data" refers to a means of collecting emotional data from the user's voice, facial expressions, and text, and transmitting it to a server.

[1210] This invention provides a system to support decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes user emotions. The system of this invention includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and emotion data acquisition means. Its specific form is shown below.

[1211] 1. Data acquisition methods

[1212] The server retrieves the user's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. For example, the server uses an API to collect this data.

[1213] The device monitors the user's real-time actions and behaviors, including the location and frequency of clicks and taps, and sends this data to a server. This is done using the device's sensors.

[1214] As part of the initial setup, users enter their values ​​and priorities into the system via web forms or in-app questionnaires. This input data is sent to the server via the device and stored in a database.

[1215] The emotion engine acquires emotional data from the user's voice, facial expressions, text, etc., and sends it to the server. For example, it performs facial recognition through the camera and voice analysis through the microphone.

[1216] 2. Data preprocessing means

[1217] The server cleanses all acquired data, removing inaccuracies and noise. Specifically, it removes duplicate data and imputes missing values. It also normalizes the data, converting data from different data sources into a unified format.

[1218] The terminal converts the survey results entered by the user into an appropriate data format, such as JSON or XML, and sends that data to the server.

[1219] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[1220] 3. Learning methods

[1221] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The model is trained on a training dataset, and its parameters are adjusted. The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[1222] 4. Recommended generation method

[1223] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[1224] The device displays recommendations received from the server to the user via an interface. Information is provided in real time using notifications and pop-up messages.

[1225] 5. User Feedback Methods

[1226] Users can review the provided recommendations and approve, reject, or modify them. They can also provide feedback on whether the recommendations were appropriate and whether they fit their emotional state.

[1227] The device collects user feedback in real time and sends the results to the server.

[1228] 6. Model Improvement Methods

[1229] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[1230] Specific example

[1231] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[1232] The server retrieves historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[1233] The emotion engine collects emotional data from the leader's voice and facial expressions.

[1234] The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[1235] The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[1236] The server recommends high-risk projects when you're feeling exhilarated, and stable projects when you're feeling stressed.

[1237] Leaders review, approve, or revise recommendations, and also provide emotional feedback.

[1238] The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1239] This enables the system of the present invention to support user decision-making and provide personalized recommendations that take emotions into consideration. Furthermore, this improves user productivity and efficiency and contributes to stress reduction.

[1240] Example of a prompt

[1241] The system generates an answer to the question, "Which project should we allocate resources to next month?" by considering the success rate of recent projects, resource utilization, and leader emotional data (emotional state, stress level, etc.).

[1242] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1243] Step 1:

[1244] Data collection:

[1245] The server retrieves users' past behavioral data and decision-making history from the database using multiple APIs. This includes purchase history, schedules, and message exchanges.

[1246] Input: User ID

[1247] Output: User data such as purchase history, schedule, and message history.

[1248] The device monitors the user's real-time actions (location and frequency of clicks and taps) and behavior, and periodically sends this data to the server. Activity data is captured using the device's sensors.

[1249] Input: User operation data

[1250] Output: Real-time operation data

[1251] As part of the initial setup, users enter their values ​​and priorities through web forms or in-app questionnaires. This data is transmitted to the server via the device and stored in a database.

[1252] Input: Survey data

[1253] Output: Values ​​and priority data

[1254] The emotion engine acquires emotional data from the user's voice, facial expressions, and text, and sends it to the server. For example, it recognizes facial expressions through the camera and performs voice analysis through the microphone.

[1255] Input: Voice, facial expressions, text data

[1256] Output: Sentiment data

[1257] Step 2:

[1258] Data preprocessing:

[1259] The server cleanses the acquired data, removing inaccuracies and noise. Specific tasks include deleting duplicate data and imputing missing values. It also normalizes the data, converting data from different sources into a unified format.

[1260] Input: Raw data (purchase history, schedule, message history, real-time operation data, sentiment data)

[1261] Output: Cleansed and normalized data

[1262] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[1263] Input: Survey results

[1264] Output: Survey data in JSON or XML format

[1265] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[1266] Input: Sentiment data

[1267] Output: Preprocessed sentiment data

[1268] Step 3:

[1269] Model learning:

[1270] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process. The model is trained on a training dataset and its parameters are adjusted.

[1271] Input: Cleansed and normalized data

[1272] Output: Trained model

[1273] The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[1274] Input: Training data

[1275] Output: High-precision model

[1276] Step 4:

[1277] Recommended generation:

[1278] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[1279] Input: Trained model, current situation data, sentiment data

[1280] Output: Recommendations

[1281] The terminal notifies the user of recommendations received from the server and displays them on the interface. Information is provided in real time using notifications and pop-up messages.

[1282] Input: Recommendations

[1283] Output: Display on the user screen

[1284] Step 5:

[1285] Collecting user feedback:

[1286] Users review the provided recommendations and can approve, reject, or modify them. They also provide feedback on whether the recommendations are appropriate and whether they align with their emotional state.

[1287] Input: User feedback on recommendations

[1288] Output: Revised recommendations, feedback data

[1289] The device collects user feedback in real time and sends the results to the server.

[1290] Input: User feedback data

[1291] Output: Feedback data sent to the server

[1292] Step 6:

[1293] Model improvements:

[1294] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[1295] Input: Feedback data, new behavioral data, emotional data

[1296] Output: Improved model

[1297] This allows the system to support user decision-making, provide personalized recommendations that take emotions into account, and improve user productivity and efficiency, as well as reduce stress.

[1298] (Application Example 2)

[1299] 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."

[1300] In modern factories and production sites, decision-making regarding robot operation and maintenance is crucial. However, many systems fail to consider the emotions and stress levels of managers, leading to inefficient resource allocation and increased robot failure risks. Therefore, decision support systems that take managers' emotions into account need to provide more optimal resource allocation and maintenance schedules.

[1301] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means. This makes it possible to grasp the emotional state of the administrator in real time and recommend appropriate resource allocation and maintenance schedules based on that.

[1302] "Data collection means" refers to methods for acquiring users' past behavioral data and decision-making history, as well as real-time operation data and emotional data.

[1303] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, removing inaccurate data and noise, and converting it into a format suitable for analysis.

[1304] "Learning methods" refer to techniques that use deep learning models or machine learning algorithms to model user decision-making patterns based on user behavior data, decision history, values ​​data, sentiment data, etc.

[1305] A "recommendation generation method" is a means of generating appropriate recommendations based on the current user situation and sentiment data, using a trained model.

[1306] A "user feedback mechanism" refers to a means of collecting feedback from users, such as modifications or approvals of recommendations provided by users.

[1307] "Model improvement methods" refer to techniques for retraining a model using user feedback and new behavioral and sentiment data to improve the accuracy of recommendations.

[1308] "Emotion recognition means" refers to methods for acquiring and analyzing emotional data from a user's voice, facial expressions, text, etc.

[1309] A "real-time decision support system" is a means of providing appropriate recommendations immediately based on data and sentiment data collected in real time, thereby supporting the user's decision-making.

[1310] The embodiments for carrying out this invention will be described below. The factory robot management assistant system includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means.

[1311] System Configuration

[1312] Data acquisition methods

[1313] In this system, the server retrieves the administrator's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. The terminal monitors real-time operation data and sends it to the server. The emotion engine, used as an emotion recognition tool, retrieves emotion data from the administrator's voice, facial expressions, text, etc., and sends it to the server.

[1314] Data preprocessing means

[1315] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis. On the terminal side, the survey results entered by the user are converted into an appropriate data format (e.g., JSON or XML) and sent to the server. Sentiment data is also preprocessed and sent to the server.

[1316] Learning methods

[1317] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The learned model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[1318] Recommended generation method

[1319] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it recommends high-risk projects when the administrator is emotionally uplifted, and stable projects when they are stressed. The terminal notifies the user of the recommendations received from the server and displays them in the interface.

[1320] User Feedback Methods

[1321] Users review the provided recommendations and provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide emotional feedback. The device collects user feedback in real time and sends the data to the server.

[1322] Model Improvement Means

[1323] The server retrains its model using user feedback and new behavioral and sentiment data to improve the accuracy of its recommendations.

[1324] Hardware and software to be used

[1325] Hardware: Computer, camera, and microphone built into the factory robot.

[1326] Software: Python, scikit-learn, emotion_recognition library

[1327] Processing Overview

[1328] 1. Data collection:

[1329] The server retrieves the administrator's past behavioral data and decision-making history from a database, while the terminal monitors real-time operation data. The emotion engine collects emotional data from the administrator's voice and facial expressions.

[1330] 2. Data preprocessing:

[1331] The acquired data is cleansed and normalized by the server. Data such as survey results are also converted into an appropriate format.

[1332] 3. Learning:

[1333] The server uses a deep learning model to learn from the administrator's behavioral and emotional data.

[1334] 4. Recommended generation:

[1335] Based on real-time data and pre-trained models, it generates recommendations tailored to the current administrator situation.

[1336] 5. Gathering feedback:

[1337] User feedback is sent to the server via the device.

[1338] 6. Model Improvement:

[1339] The server retrains the model based on the feedback, improving the accuracy of future recommendations.

[1340] Specific example

[1341] For example, if a factory robot manager is under a very high level of stress, this management assistant app will recommend reducing the high-risk tasks performed by the robots. Conversely, if the manager is emotionally stable, it will prioritize efficiency and make recommendations that include high-risk tasks.

[1342] Example of a prompt

[1343] "Based on the following data, please generate the optimal work schedule for factory robot management. The data is as follows:"

[1344] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1345] Step 1:

[1346] The server retrieves the administrator's past behavioral data and decision history from the database. This includes information such as purchase history, schedules, and message exchanges. The input is past behavioral data and decision history from the database, and the output is a dataset containing this data. Specifically, it uses SQL queries to retrieve the necessary information from the database.

[1347] Step 2:

[1348] The terminal monitors real-time operation and emotion data and sends it to the server. Inputs are administrator operations and voice and facial expression data collected by the emotion engine, while output is real-time data sent to the server. Specifically, emotion recognition software is used to extract emotion information from voice and facial expressions.

[1349] Step 3:

[1350] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data. The input is the data obtained in steps 1 and 2, and the output is the cleansed and normalized dataset. Specifically, it uses a data cleaning tool to remove incomplete data and noise, and a standard scaler to normalize the data.

[1351] Step 4:

[1352] The server performs deep learning using recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and sentiment data. The input is a cleansed and normalized dataset, and the output is the trained recommendation model. Specifically, the model is trained using deep learning frameworks such as PyTorch and TensorFlow.

[1353] Step 5:

[1354] The server uses a pre-trained model to generate recommendations based on the current user's situation and sentiment. The input is real-time interaction and sentiment data, and the output is recommendations. Specifically, it inputs real-time data into the pre-trained model and generates recommendations.

[1355] Step 6:

[1356] The terminal notifies the user of recommendations received from the server and displays them on the interface. The input is the recommendations from the server, and the output is the user's feedback. Specifically, it displays the recommendations using the user interface (UI) and receives feedback from the user.

[1357] Step 7:

[1358] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The input is the recommendations displayed on the device, and the output is the feedback sent to the device. Specifically, the user approves, rejects, or modifies the recommendations and inputs the result into the device.

[1359] Step 8:

[1360] The server retrains the model using user feedback and new behavioral and sentiment data. The input is user feedback data and new data, and the output is the updated trained model. Specifically, it retrains the model using feedback data to improve accuracy.

[1361] 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.

[1362] 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.

[1363] 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.

[1364] [Fourth Embodiment]

[1365] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1366] 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.

[1367] 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).

[1368] 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.

[1369] 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.

[1370] 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).

[1371] 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.

[1372] 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.

[1373] 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.

[1374] 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.

[1375] 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.

[1376] 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.

[1377] 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".

[1378] This invention is a system that supports decision-making for busy business leaders, managers, and experts, and its specific form is shown below.

[1379] System Configuration

[1380] The system of this invention mainly consists of the following elements:

[1381] 1. Data acquisition methods

[1382] 2. Data preprocessing means

[1383] 3. Learning methods

[1384] 4. Recommended generation method

[1385] 5. User Feedback Methods

[1386] 6. Model Improvement Methods

[1387] System operation

[1388] Data acquisition methods

[1389] The server retrieves the user's past behavioral data and decision-making history from a database. This includes purchase history, schedules, and message exchanges.

[1390] The terminal monitors user operations and actions in real time and sends this information to the server.

[1391] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This information is also sent to the server and stored in the database.

[1392] Data preprocessing means

[1393] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analysis.

[1394] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[1395] Learning methods

[1396] The server learns the user's decision-making process using deep learning models (e.g., recurrent neural networks or transformer models). This allows it to extract individual user patterns and incorporate them into the model.

[1397] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[1398] Recommended generation method

[1399] The server uses a pre-trained model to generate user-specific recommendations. For example, when a user is deciding on project resource allocation, it suggests the best option based on past success rates and cost-effectiveness.

[1400] The terminal notifies the user of recommendations from the server and displays them on the interface.

[1401] User Feedback Methods

[1402] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, they also send the results to the server.

[1403] The device collects user feedback in real time and sends it to the server.

[1404] Model Improvement Means

[1405] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[1406] Specific example

[1407] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[1408] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[1409] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[1410] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[1411] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[1412] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[1413] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1414] In this way, the system of the present invention supports the user's decision-making process and provides optimized decision-making. This enables personalized support based on individual values, improving the user's productivity and efficiency.

[1415] The following describes the processing flow.

[1416] ---

[1417] Step 1:

[1418] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[1419] The device monitors the user's real-time operations and actions and sends that data to the server.

[1420] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server.

[1421] Step 2:

[1422] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[1423] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[1424] Step 3:

[1425] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data and values. This allows it to model the user's decision-making process.

[1426] The trained model reflects the user's characteristics and past decision-making patterns.

[1427] Step 4:

[1428] The server uses a pre-trained model to generate recommendations based on the user's current situation. For example, it might recommend which project the user should allocate resources to next.

[1429] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[1430] Step 5:

[1431] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[1432] The device collects user feedback in real time and sends that data to the server.

[1433] Step 6:

[1434] The server uses user feedback and new behavioral data to retrain the model, aiming for personalization and improved accuracy.

[1435] This will improve the quality of future recommendations and more effectively support user decision-making.

[1436] ---

[1437] The above describes the specific processing flow in the system of the present invention. This enables efficient and consistent support for the user's decision-making process.

[1438] (Example 1)

[1439] 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".

[1440] Traditional decision support systems have faced challenges in effectively assisting busy business leaders and managers with their decision-making. In particular, their inability to provide personalized recommendations that reflect individual user values ​​and priorities often leads to a decline in recommendation quality and reduced decision-making efficiency. Furthermore, the difficulty in effectively incorporating user feedback to improve model accuracy results in a gradual deterioration of recommendation quality.

[1441] 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.

[1442] In this invention, the server includes data collection means for acquiring the user's past behavioral data and decision-making history, data preprocessing means for cleansing and normalizing the acquired data, learning means using a deep learning model, recommendation generation means for generating recommendations, user feedback means for receiving user feedback, and model improvement means for improving the accuracy of the model. This makes it possible to provide personalized recommendations based on the user's individual behavior and values, and to continuously improve the quality of recommendations and the efficiency of decision-making.

[1443] "Data collection methods" refer to means of acquiring users' past behavioral data and decision-making history.

[1444] "Data preprocessing means" refers to methods for cleansing and normalizing acquired data.

[1445] A "learning method" is a means of learning user decision-making patterns using a deep learning model.

[1446] A "recommendation generation method" is a method for generating recommendations for users using a pre-trained model.

[1447] A "user feedback method" is a means of receiving user feedback.

[1448] "Model improvement methods" are means of improving the accuracy of a model based on user feedback.

[1449] This invention is a system that supports decision-making for busy business leaders, managers, and professionals. This system primarily consists of the following elements:

[1450] 1. Data acquisition methods

[1451] 2. Data preprocessing means

[1452] 3. Learning methods

[1453] 4. Recommended generation method

[1454] 5. User Feedback Methods

[1455] 6. Model Improvement Methods

[1456] Data acquisition methods

[1457] The server retrieves data on the user's past behavior and decision-making history from the database. For example, it collects data such as purchase history, schedules, and message exchanges. This provides the foundational data needed to understand the user's decision-making patterns.

[1458] The device monitors user actions and behaviors in real time and sends that information to the server. For example, it monitors actions such as a user adding an event to their calendar in real time and sends that data to the server.

[1459] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[1460] Data preprocessing means

[1461] The server cleanses the acquired data, removing inconsistent data. Furthermore, it normalizes the data and converts it into a format suitable for analysis. For example, it might unify date data in different formats.

[1462] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. This process ensures data consistency.

[1463] Learning methods

[1464] The server learns user decision-making patterns using deep learning models (e.g., recurrent neural networks or transformer models). Specifically, data is divided into batches and fed into the model, and learning is performed epoch by epoch.

[1465] The training process is periodically rerun in batches, and the model is continuously updated with new data. This process allows the model to make recommendations that reflect the latest data.

[1466] Recommended generation method

[1467] The server uses a pre-trained model to generate personalized recommendations based on the user's situation. For example, when a business leader is deciding which projects to allocate resources to next month, the server suggests the best options based on past success rates and cost-effectiveness.

[1468] The terminal notifies the user of recommendations from the server and displays them on the interface. Notifications are made in real time, allowing the user to check them immediately.

[1469] User Feedback Methods

[1470] Users review the provided recommendations and make modifications or provide feedback as needed. If they approve, reject, or modify the recommendations, the results are also sent to the server.

[1471] The device collects user feedback in real time and sends it to the server. The collected data is immediately reflected in the database.

[1472] Model Improvement Means

[1473] The server improves the model's accuracy using user feedback and newly collected data. By analyzing the feedback data and optimizing the model's loss function, the accuracy of the next recommendation improves.

[1474] Specific examples and prompt statements

[1475] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[1476] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[1477] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server.

[1478] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns.

[1479] 4. Recommendation generation method: The server generates recommendations for the optimal project resource allocation for the following month and notifies the leader via the terminal.

[1480] 5. User Feedback Method: Leaders review, approve, or revise recommendations. Feedback is sent to the server via the device.

[1481] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1482] Examples of prompts for a generative AI model:

[1483] "Based on past project data, please recommend the optimal resource allocation for the next month. Consider variables such as success rate, cost, and time."

[1484] By using the above system, it becomes possible to support the user's decision-making process and provide personalized recommendations, thereby improving productivity and efficiency.

[1485] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1486] Step 1: Data Collection

[1487] The server retrieves the user's past behavioral data and decision-making history from the database. Specifically, it issues queries to collect data such as purchase history, schedules, and message exchanges. The input is the query results from the database, and the output is the set of collected data.

[1488] The device monitors user actions and behaviors in real time and sends that information to the server. Specifically, it captures actions such as a user adding an event to their calendar and sends that data to the server. The input is the user's real-time actions, and the output is the monitored data.

[1489] As part of the initial setup in the form of a questionnaire, users enter their values ​​and priorities. This data is also sent to the server and stored in the database. The input is the user's questionnaire responses, and the output is the data sent to the server.

[1490] Step 2: Data Preprocessing

[1491] The server cleanses the acquired data and removes inconsistent data. Specifically, it imputes missing values ​​and removes outliers. The input is the collected raw data, and the output is the cleansed data.

[1492] The server further normalizes the data and standardizes its format. For example, it unifies date data in different formats. The input is cleansed data, and the output is normalized data.

[1493] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends it to the server. The input is the user's survey data, and the output is the converted data.

[1494] Step 3: Model Training

[1495] The server trains a deep learning model using normalized data. Specifically, it batch processes the data using recurrent neural networks or transformer models and performs training in epoch units. The input is a normalized dataset, and the output is the trained model.

[1496] The training process is periodically rerun in batches, and the model is continuously updated using newly collected data. The input is the new data, and the output is the updated model.

[1497] Step 4: Recommended Generation

[1498] The server uses a trained model to generate user-specific recommendations. For example, it might suggest the optimal project resource allocation for a business leader for the following month. The input is the trained model and current user data, and the output is the generated recommendations.

[1499] The terminal notifies the user of recommendations received from the server and displays them on the interface. Specifically, it presents recommendations to the user through pop-up notifications. The input is the recommendations, and the output is the notification to the user.

[1500] Step 5: User Feedback

[1501] Users review the provided recommendations and make corrections or provide feedback as needed. For example, they may be given options to accept, reject, or modify the recommendations. The input is the user's feedback, and the output is the modified or accepted data.

[1502] The device collects user feedback in real time and sends it to the server. The input is the user's real-time feedback, and the output is the data sent to the server.

[1503] Step 6: Model Improvement

[1504] The server improves the model's accuracy using user feedback and newly collected data. Specifically, it analyzes the feedback data and optimizes the model's loss function. The input is the feedback data and newly collected data, and the output is the updated model.

[1505] By following these steps, the system can continue to provide recommendations optimized for the user's situation.

[1506] (Application Example 1)

[1507] 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".

[1508] In conventional factory robot operations, it was difficult to efficiently manage the operating data, maintenance records, and production plans of individual robots, making it challenging to determine the optimal operating schedule and maintenance timing. As a result, production efficiency decreased and costs increased due to unnecessary maintenance. The present invention aims to solve these problems and provide a new system for optimizing robot operations within a factory.

[1509] 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.

[1510] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, means for acquiring and analyzing data to optimize the operational efficiency of robots in a factory, means for recommending an optimal operational schedule and maintenance timing, and means for displaying the recommended operational schedule and maintenance timing. This enables efficient operational management and appropriate maintenance of factory robots.

[1511] "Data collection means" refers to means for acquiring information such as users' past behavioral data, decision-making history, robot operation data within the factory, maintenance records, and production plans.

[1512] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, and further converting it into a format suitable for analyzing the operational efficiency of robots within a factory.

[1513] "Learning methods" refer to techniques that improve the accuracy of recommendations by using deep learning models based on user and robot behavior data.

[1514] A "recommendation generation method" is a means of generating recommendations to optimize the next task or maintenance timing of a factory robot using a pre-trained model.

[1515] A "user feedback mechanism" is a means of collecting user feedback and modifications to generated recommendations and incorporating them into the system.

[1516] "Model improvement methods" refer to techniques for continuously improving the accuracy of a learning model using user feedback and new data.

[1517] "Means of acquiring and analyzing data to optimize operational efficiency" refers to methods of collecting data on robots within a factory and analyzing it to achieve optimal operational efficiency.

[1518] "Means for recommending optimal operating schedules and maintenance timings" refers to methods for proposing optimal operating schedules and maintenance timings for robots to users based on collected and analyzed data.

[1519] "Means for displaying recommended operational schedules and maintenance timings" refers to a means of visually communicating the generated recommendations by displaying them on the user's device.

[1520] This invention is a system for optimizing the operational efficiency of robots in a factory, and its specific form is shown below.

[1521] System Configuration

[1522] The system of the present invention consists of the following elements:

[1523] 1. Data acquisition methods

[1524] 2. Data preprocessing means

[1525] 3. Learning methods

[1526] 4. Recommended generation method

[1527] 5. User Feedback Methods

[1528] 6. Model Improvement Methods

[1529] 7. Means for acquiring and analyzing data to optimize operational efficiency

[1530] 8. Means for recommending optimal operating schedules and maintenance timings.

[1531] 9. Means for displaying recommended operational schedules and maintenance timings.

[1532] System operation

[1533] Data acquisition methods

[1534] The server retrieves operational data, maintenance records, and production plans from robots within the factory via an API.

[1535] Users provide past behavioral data and decision-making history, which is also sent to the server and stored in the database.

[1536] Data preprocessing means

[1537] The server cleanses the acquired data, removing inconsistent data. It also normalizes the data and converts it into a format suitable for analyzing operational efficiency within the factory.

[1538] The terminal converts the data obtained from the user into an appropriate data format (e.g., JSON or XML) and sends it to the server.

[1539] Learning methods

[1540] The server uses deep learning models (e.g., recurrent neural networks or transformer models) to learn user and robot behavior data and incorporate it into the model.

[1541] The training process is periodically rerun in batches, and the model is continuously updated with new data.

[1542] Recommended generation method

[1543] The server uses pre-trained models to recommend optimizations for the next tasks and maintenance timings of factory robots.

[1544] The terminal notifies the user of recommendations from the server and displays them on the interface.

[1545] User Feedback Methods

[1546] Users review the provided recommendations and make corrections or provide feedback. This feedback is sent to the server.

[1547] The device collects user feedback in real time and sends it to the server.

[1548] Model Improvement Means

[1549] The server uses user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided.

[1550] Specific example

[1551] As a concrete example, consider a scenario where a factory manager decides which robots should be assigned which tasks in the following month.

[1552] 1. Data collection method: The server acquires historical data such as the operation data, success rate, maintenance records, and production plans of factory robots.

[1553] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the data into a format for analyzing the robot's operational efficiency and sends it to the server.

[1554] 3. Learning method: The server uses a deep learning model (e.g., Keras / TensorFlow) to learn the robot's operational patterns.

[1555] 4. Recommendation generation method: The server generates recommendations for the optimal task assignment for the following month and notifies the manager via the terminal.

[1556] 5. User Feedback Method: Managers review, approve, or revise recommendations. Feedback is sent to the server via the terminal.

[1557] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1558] This optimizes the operational schedule and maintenance timing of robots within the factory, improving productivity and efficiency.

[1559] Example of a prompt

[1560] "Please retrieve robot operation data from the factory data API."

[1561] Please normalize your data using Pandas and Scikit-learn.

[1562] "Load the saved Keras model and perform inference."

[1563] "Generate recommendations based on the model's inference results."

[1564] "Please display the generated recommendations in JSON format."

[1565] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1566] Step 1:

[1567] The server retrieves information such as robot operation data, maintenance records, and production plans from robots within the factory via an API. This includes each robot's work history, current status, and future work schedule. Input is raw data retrieved from the API, and output is canonical data stored in the database.

[1568] Step 2:

[1569] The server cleanses the acquired raw data and removes inconsistent data. Next, it normalizes the data and converts it into a format suitable for analyzing the operational efficiency of factory robots. Data processing libraries such as Pandas and Scikit-learn are used for this. The input is the raw data acquired in step 1, and the output is the cleansed and normalized data.

[1570] Step 3:

[1571] The server uses a deep learning model (e.g., Keras / TensorFlow) to train on normalized data. The training process involves pattern extraction and model training using recurrent neural networks or transformer models based on acquired historical data. The input is the data preprocessed in step 2, and the output is the trained model.

[1572] Step 4:

[1573] The server uses a pre-trained model to perform inference and recommend optimizations for the next task and maintenance timing of factory robots. This involves converting the generated recommendations into JSON format and sending them to the terminal. The input is the pre-trained model and current data, and the output is the recommendations in JSON format.

[1574] Step 5:

[1575] The terminal notifies the user of recommendations received from the server and displays them on the interface. This allows the user to review the recommendations and provide feedback or make corrections as needed. The input is the recommendations received from the server in JSON format, and the output is the recommendations displayed in a format usable by the user.

[1576] Step 6:

[1577] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The feedback is sent to the server via the device. The input is the recommendations displayed on the device, and the output is the feedback sent to the server.

[1578] Step 7:

[1579] The server uses received user feedback and newly collected data to personalize and improve the accuracy of the model. This continuously improves the quality of the recommendations provided. The input is user feedback and new data, and the output is the updated trained model.

[1580] 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.

[1581] This invention is a system that supports decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes the user's emotions. Its specific form is described below.

[1582] System Configuration

[1583] The system of the present invention consists of the following elements:

[1584] 1. Data acquisition methods

[1585] 2. Data preprocessing means

[1586] 3. Learning methods

[1587] 4. Recommended generation method

[1588] 5. User Feedback Methods

[1589] 6. Model Improvement Methods

[1590] 7. Emotional Engine

[1591] System operation

[1592] Data acquisition methods

[1593] The server retrieves the user's past behavioral data and decision-making history from the database. This includes information such as purchase history, schedule, and message exchanges.

[1594] The device monitors the user's real-time operations and actions and sends that data to the server.

[1595] As part of the initial setup, users enter their values ​​and priorities in a questionnaire format. This input data is also sent to the server and stored in the database.

[1596] The emotion engine acquires emotion data from the user's voice, facial expressions, text, etc., and sends it to the server.

[1597] Data preprocessing means

[1598] The server cleanses all acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis.

[1599] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[1600] The emotion data acquired by the emotion engine is also appropriately preprocessed before being sent to the server.

[1601] Learning methods

[1602] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process.

[1603] The trained model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[1604] Recommended generation method

[1605] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it might recommend which project to allocate resources to next, taking into account whether the user is feeling uplifted or stressed.

[1606] The terminal notifies the user of recommendations received from the server and displays them on the interface.

[1607] User Feedback Methods

[1608] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide feedback regarding their own feelings.

[1609] The device collects user feedback in real time and sends that data to the server.

[1610] Model Improvement Means

[1611] The server uses user feedback, new behavioral data, and sentiment data to retrain the model, aiming for personalization and improved accuracy.

[1612] This will improve the quality of future recommendations and more effectively support user decision-making.

[1613] Specific example

[1614] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[1615] 1. Data Collection Methods: The server acquires historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data. The emotion engine collects emotional data from the leader's voice and facial expressions.

[1616] 2. Data preprocessing means: The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[1617] 3. Learning method: The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[1618] 4. Recommended generation method: The server recommends high-risk projects when the user is feeling exhilarated, and stable projects when the user is feeling stressed.

[1619] 5. User Feedback Methods: Leaders review, approve, or revise recommendations. They also provide emotional feedback.

[1620] 6. Model Improvement Method: The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1621] This enables the system of the present invention to support the user's decision-making process and provide personalized recommendations that take emotions into consideration. This, in turn, improves user productivity and efficiency and contributes to stress reduction.

[1622] The following describes the processing flow.

[1623] ---

[1624] Step 1:

[1625] The server retrieves the user's past behavioral data and decision-making history from the database. This includes purchase history, project evaluations, and the results of past decisions.

[1626] The device monitors the user's real-time actions and behavior and sends that data to the server. For example, it might track which web pages the user is browsing or which apps they are using.

[1627] The emotion engine acquires emotional data from the user's voice, facial expressions, and text. For example, it analyzes the user's stress level and happiness level from the tone of their speech and facial expressions.

[1628] Step 2:

[1629] The server cleanses all acquired data, removing inaccurate data and noise. For example, it removes duplicate and inconsistent data.

[1630] The server performs normalization and transforms the data into a format suitable for analysis. For example, it scales numerical data or encodes categorical data.

[1631] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server. For example, it extracts text data from the survey form and converts it into JSON format.

[1632] The emotion engine preprocesses emotion data and sends it to the server. For example, it converts audio data to text and assigns emotion labels to it.

[1633] Step 3:

[1634] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process.

[1635] The training process is periodically rerun in batches, and the model is continuously updated with new data. For example, the model is retrained every night with new data.

[1636] Step 4:

[1637] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is feeling stressed, it will recommend safer options.

[1638] The device notifies the user of recommendations received from the server and displays them on the interface. For example, it might use smartphone push notifications or notifications from wearable devices.

[1639] Step 5:

[1640] Users review the provided recommendations and make modifications or provide feedback as needed. For example, they can accept, reject, or partially modify the recommendations.

[1641] The device collects user feedback in real time and sends that data to the server. For example, when a user presses the "Approve" button in a push notification, that selection is recorded and sent.

[1642] Step 6:

[1643] The server uses user feedback and new behavioral and sentiment data to retrain its model, aiming for personalization and improved accuracy.

[1644] This improves the quality of future recommendations and more effectively supports user decision-making. For example, new recommendations can be readjusted based on the user's previous acceptance of a recommendation and their subsequent emotional response.

[1645] ---

[1646] The above describes the specific processing flow in the system of the present invention that incorporates an emotion engine. This enables the user's decision-making process to be supported efficiently and consistently while taking emotions into consideration.

[1647] (Example 2)

[1648] 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".

[1649] Busy business leaders, managers, and professionals face a challenge in making quick and accurate decisions, as they are often influenced by emotions. This can lead to stress and decreased productivity. Therefore, there is a need for systems that support decision-making by providing personalized recommendations while considering the user's emotional state.

[1650] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1651] In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and sentiment data acquisition means. This makes it possible to provide personalized and sentiment-responsive recommendations based on user behavior data and sentiment data.

[1652] "Data collection means" refers to means for acquiring users' past behavioral data and decision-making history and transmitting it to a server.

[1653] "Data preprocessing means" refers to means for cleansing and normalizing acquired data and converting it into a format suitable for analysis.

[1654] "Learning methods" refer to techniques for performing deep learning using recurrent neural networks (RNNs) or transformer models based on user behavior data, values, and emotional data.

[1655] A "recommendation generation method" is a means of generating recommendations based on the current user's situation and emotions, using a pre-trained model.

[1656] A "user feedback mechanism" is a means for users to provide feedback on the recommendations they have been given, to collect that feedback in real time, and to send it to the server.

[1657] "Model improvement methods" refer to techniques for retraining a model using user feedback, new behavioral data, and sentiment data, with the aim of improving personalization and accuracy.

[1658] "Means for acquiring emotional data" refers to a means of collecting emotional data from the user's voice, facial expressions, and text, and transmitting it to a server.

[1659] This invention provides a system to support decision-making for busy business leaders, managers, and professionals, and incorporates an emotion engine that recognizes user emotions. The system of this invention includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, and emotion data acquisition means. Its specific form is shown below.

[1660] 1. Data acquisition methods

[1661] The server retrieves the user's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. For example, the server uses an API to collect this data.

[1662] The device monitors the user's real-time actions and behaviors, including the location and frequency of clicks and taps, and sends this data to a server. This is done using the device's sensors.

[1663] As part of the initial setup, users enter their values ​​and priorities into the system via web forms or in-app questionnaires. This input data is sent to the server via the device and stored in a database.

[1664] The emotion engine acquires emotional data from the user's voice, facial expressions, text, etc., and sends it to the server. For example, it performs facial recognition through the camera and voice analysis through the microphone.

[1665] 2. Data preprocessing means

[1666] The server cleanses all acquired data, removing inaccuracies and noise. Specifically, it removes duplicate data and imputes missing values. It also normalizes the data, converting data from different data sources into a unified format.

[1667] The terminal converts the survey results entered by the user into an appropriate data format, such as JSON or XML, and sends that data to the server.

[1668] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[1669] 3. Learning methods

[1670] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The model is trained on a training dataset, and its parameters are adjusted. The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[1671] 4. Recommended generation method

[1672] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[1673] The device displays recommendations received from the server to the user via an interface. Information is provided in real time using notifications and pop-up messages.

[1674] 5. User Feedback Methods

[1675] Users can review the provided recommendations and approve, reject, or modify them. They can also provide feedback on whether the recommendations were appropriate and whether they fit their emotional state.

[1676] The device collects user feedback in real time and sends the results to the server.

[1677] 6. Model Improvement Methods

[1678] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[1679] Specific example

[1680] Specific example: When a business leader decides which projects to allocate resources to in the following month.

[1681] The server retrieves historical project data, success rates, costs, and time, while the terminal monitors the leader's schedule and past evaluation data.

[1682] The emotion engine collects emotional data from the leader's voice and facial expressions.

[1683] The server cleanses and normalizes the acquired data. The terminal converts the reader's priorities into an appropriate format and sends them to the server. Sentiment data is also preprocessed and sent to the server.

[1684] The server uses a deep learning model to learn the leader's decision-making patterns and sentiment data.

[1685] The server recommends high-risk projects when you're feeling exhilarated, and stable projects when you're feeling stressed.

[1686] Leaders review, approve, or revise recommendations, and also provide emotional feedback.

[1687] The server receives feedback and updates the model to improve the accuracy of the next recommendation.

[1688] This enables the system of the present invention to support user decision-making and provide personalized recommendations that take emotions into consideration. Furthermore, this improves user productivity and efficiency and contributes to stress reduction.

[1689] Example of a prompt

[1690] The system generates an answer to the question, "Which project should we allocate resources to next month?" by considering the success rate of recent projects, resource utilization, and leader emotional data (emotional state, stress level, etc.).

[1691] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1692] Step 1:

[1693] Data collection:

[1694] The server retrieves users' past behavioral data and decision-making history from the database using multiple APIs. This includes purchase history, schedules, and message exchanges.

[1695] Input: User ID

[1696] Output: User data such as purchase history, schedule, and message history.

[1697] The device monitors the user's real-time actions (location and frequency of clicks and taps) and behavior, and periodically sends this data to the server. Activity data is captured using the device's sensors.

[1698] Input: User operation data

[1699] Output: Real-time operation data

[1700] As part of the initial setup, users enter their values ​​and priorities through web forms or in-app questionnaires. This data is transmitted to the server via the device and stored in a database.

[1701] Input: Survey data

[1702] Output: Values ​​and priority data

[1703] The emotion engine acquires emotional data from the user's voice, facial expressions, and text, and sends it to the server. For example, it recognizes facial expressions through the camera and performs voice analysis through the microphone.

[1704] Input: Voice, facial expressions, text data

[1705] Output: Sentiment data

[1706] Step 2:

[1707] Data preprocessing:

[1708] The server cleanses the acquired data, removing inaccuracies and noise. Specific tasks include deleting duplicate data and imputing missing values. It also normalizes the data, converting data from different sources into a unified format.

[1709] Input: Raw data (purchase history, schedule, message history, real-time operation data, sentiment data)

[1710] Output: Cleansed and normalized data

[1711] The terminal converts the survey results entered by the user into an appropriate data format (e.g., JSON or XML) and sends that data to the server.

[1712] Input: Survey results

[1713] Output: Survey data in JSON or XML format

[1714] The emotion data acquired by the emotion engine is also preprocessed and sent to the server. Based on the emotion recognition algorithm, noise reduction and other processes are performed.

[1715] Input: Sentiment data

[1716] Output: Preprocessed sentiment data

[1717] Step 3:

[1718] Model learning:

[1719] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows it to model the user's decision-making process. The model is trained on a training dataset and its parameters are adjusted.

[1720] Input: Cleansed and normalized data

[1721] Output: Trained model

[1722] The trained model reflects the user's characteristics, past decision-making patterns, and emotional changes.

[1723] Input: Training data

[1724] Output: High-precision model

[1725] Step 4:

[1726] Recommended generation:

[1727] The server uses a pre-trained model to generate recommendations based on the user's current situation and emotions. For example, if the user is stressed, it will suggest low-risk projects; if they are feeling exhilarated, it will suggest challenging projects.

[1728] Input: Trained model, current situation data, sentiment data

[1729] Output: Recommendations

[1730] The terminal notifies the user of recommendations received from the server and displays them on the interface. Information is provided in real time using notifications and pop-up messages.

[1731] Input: Recommendations

[1732] Output: Display on the user screen

[1733] Step 5:

[1734] Collecting user feedback:

[1735] Users review the provided recommendations and can approve, reject, or modify them. They also provide feedback on whether the recommendations are appropriate and whether they align with their emotional state.

[1736] Input: User feedback on recommendations

[1737] Output: Revised recommendations, feedback data

[1738] The device collects user feedback in real time and sends the results to the server.

[1739] Input: User feedback data

[1740] Output: Feedback data sent to the server

[1741] Step 6:

[1742] Model improvements:

[1743] The server retrains the model using user feedback, new behavioral data, and sentiment data. This allows the model to become more accurate and provide more personalized recommendations in subsequent recommendation generation.

[1744] Input: Feedback data, new behavioral data, emotional data

[1745] Output: Improved model

[1746] This allows the system to support user decision-making, provide personalized recommendations that take emotions into account, and improve user productivity and efficiency, as well as reduce stress.

[1747] (Application Example 2)

[1748] 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".

[1749] In modern factories and production sites, decision-making regarding robot operation and maintenance is crucial. However, many systems fail to consider the emotions and stress levels of managers, leading to inefficient resource allocation and increased robot failure risks. Therefore, decision support systems that take managers' emotions into account need to provide more optimal resource allocation and maintenance schedules.

[1750] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means. This makes it possible to grasp the emotional state of the administrator in real time and recommend appropriate resource allocation and maintenance schedules based on that.

[1751] "Data collection means" refers to methods for acquiring users' past behavioral data and decision-making history, as well as real-time operation data and emotional data.

[1752] "Data preprocessing means" refers to means of cleansing and normalizing acquired data, removing inaccurate data and noise, and converting it into a format suitable for analysis.

[1753] "Learning methods" refer to techniques that use deep learning models or machine learning algorithms to model user decision-making patterns based on user behavior data, decision history, values ​​data, sentiment data, etc.

[1754] A "recommendation generation method" is a means of generating appropriate recommendations based on the current user situation and sentiment data, using a trained model.

[1755] A "user feedback mechanism" refers to a means of collecting feedback from users, such as modifications or approvals of recommendations provided by users.

[1756] "Model improvement methods" refer to techniques for retraining a model using user feedback and new behavioral and sentiment data to improve the accuracy of recommendations.

[1757] "Emotion recognition means" refers to methods for acquiring and analyzing emotional data from a user's voice, facial expressions, text, etc.

[1758] A "real-time decision support system" is a means of providing appropriate recommendations immediately based on data and sentiment data collected in real time, thereby supporting the user's decision-making.

[1759] The embodiments for carrying out this invention will be described below. The factory robot management assistant system includes data collection means, data preprocessing means, learning means, recommendation generation means, user feedback means, model improvement means, emotion recognition means, and real-time decision support means.

[1760] System Configuration

[1761] Data acquisition methods

[1762] In this system, the server retrieves the administrator's past behavioral data and decision-making history from a database. This includes information such as purchase history, schedules, and message exchanges. The terminal monitors real-time operation data and sends it to the server. The emotion engine, used as an emotion recognition tool, retrieves emotion data from the administrator's voice, facial expressions, text, etc., and sends it to the server.

[1763] Data preprocessing means

[1764] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data and converts it into a format suitable for analysis. On the terminal side, the survey results entered by the user are converted into an appropriate data format (e.g., JSON or XML) and sent to the server. Sentiment data is also preprocessed and sent to the server.

[1765] Learning methods

[1766] The server uses deep learning with recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and emotional data. This allows for a more precise modeling of the user's decision-making process. The learned model reflects the user's characteristics, past decision-making patterns, and even changes in their emotions.

[1767] Recommended generation method

[1768] The server uses a trained model to generate recommendations based on the user's current situation and emotions. For example, it recommends high-risk projects when the administrator is emotionally uplifted, and stable projects when they are stressed. The terminal notifies the user of the recommendations received from the server and displays them in the interface.

[1769] User Feedback Methods

[1770] Users review the provided recommendations and provide feedback as needed. For example, they can accept, reject, or partially modify recommendations. They can also provide emotional feedback. The device collects user feedback in real time and sends the data to the server.

[1771] Model Improvement Means

[1772] The server retrains its model using user feedback and new behavioral and sentiment data to improve the accuracy of its recommendations.

[1773] Hardware and software to be used

[1774] Hardware: Computer, camera, and microphone built into the factory robot.

[1775] Software: Python, scikit-learn, emotion_recognition library

[1776] Processing Overview

[1777] 1. Data collection:

[1778] The server retrieves the administrator's past behavioral data and decision-making history from a database, while the terminal monitors real-time operation data. The emotion engine collects emotional data from the administrator's voice and facial expressions.

[1779] 2. Data preprocessing:

[1780] The acquired data is cleansed and normalized by the server. Data such as survey results are also converted into an appropriate format.

[1781] 3. Learning:

[1782] The server uses a deep learning model to learn from the administrator's behavioral and emotional data.

[1783] 4. Recommended generation:

[1784] Based on real-time data and pre-trained models, it generates recommendations tailored to the current administrator situation.

[1785] 5. Gathering feedback:

[1786] User feedback is sent to the server via the device.

[1787] 6. Model Improvement:

[1788] The server retrains the model based on the feedback, improving the accuracy of future recommendations.

[1789] Specific example

[1790] For example, if a factory robot manager is under a very high level of stress, this management assistant app will recommend reducing the high-risk tasks performed by the robots. Conversely, if the manager is emotionally stable, it will prioritize efficiency and make recommendations that include high-risk tasks.

[1791] Example of a prompt

[1792] "Based on the following data, please generate the optimal work schedule for factory robot management. The data is as follows:"

[1793] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1794] Step 1:

[1795] The server retrieves the administrator's past behavioral data and decision history from the database. This includes information such as purchase history, schedules, and message exchanges. The input is past behavioral data and decision history from the database, and the output is a dataset containing this data. Specifically, it uses SQL queries to retrieve the necessary information from the database.

[1796] Step 2:

[1797] The terminal monitors real-time operation and emotion data and sends it to the server. Inputs are administrator operations and voice and facial expression data collected by the emotion engine, while output is real-time data sent to the server. Specifically, emotion recognition software is used to extract emotion information from voice and facial expressions.

[1798] Step 3:

[1799] The server cleanses the acquired data, removing inaccurate data and noise. It also normalizes the data. The input is the data obtained in steps 1 and 2, and the output is the cleansed and normalized dataset. Specifically, it uses a data cleaning tool to remove incomplete data and noise, and a standard scaler to normalize the data.

[1800] Step 4:

[1801] The server performs deep learning using recurrent neural networks (RNNs) and transformer models based on user behavior data, values, and sentiment data. The input is a cleansed and normalized dataset, and the output is the trained recommendation model. Specifically, the model is trained using deep learning frameworks such as PyTorch and TensorFlow.

[1802] Step 5:

[1803] The server uses a pre-trained model to generate recommendations based on the current user's situation and sentiment. The input is real-time interaction and sentiment data, and the output is recommendations. Specifically, it inputs real-time data into the pre-trained model and generates recommendations.

[1804] Step 6:

[1805] The terminal notifies the user of recommendations received from the server and displays them on the interface. The input is the recommendations from the server, and the output is the user's feedback. Specifically, it displays the recommendations using the user interface (UI) and receives feedback from the user.

[1806] Step 7:

[1807] The user reviews the provided recommendations and makes modifications or provides feedback as needed. The input is the recommendations displayed on the device, and the output is the feedback sent to the device. Specifically, the user approves, rejects, or modifies the recommendations and inputs the result into the device.

[1808] Step 8:

[1809] The server retrains the model using user feedback and new behavioral and sentiment data. The input is user feedback data and new data, and the output is the updated trained model. Specifically, it retrains the model using feedback data to improve accuracy.

[1810] 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.

[1811] 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.

[1812] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1813] 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.

[1814] 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.

[1815] 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.

[1816] 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.

[1817] 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.

[1818] 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."

[1819] 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.

[1820] 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.

[1821] 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.

[1822] 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.

[1823] 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.

[1824] 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.

[1825] 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.

[1826] 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.

[1827] 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.

[1828] 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.

[1829] 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.

[1830] 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.

[1831] The following is further disclosed regarding the embodiments described above.

[1832] (Claim 1)

[1833] Data collection means,

[1834] Data preprocessing means,

[1835] Learning methods,

[1836] Recommended generation method,

[1837] User feedback methods,

[1838] A system including model improvement methods.

[1839] (Claim 2)

[1840] The system according to claim 1, wherein the data collection means includes means for acquiring the user's past behavioral data and decision-making history.

[1841] (Claim 3)

[1842] The system according to claim 1, wherein the data preprocessing means includes means for cleansing and normalizing the acquired data.

[1843] (Claim 4)

[1844] The system according to claim 1, wherein the learning means includes means for learning the user's decision-making process using a deep learning model.

[1845] (Claim 5)

[1846] The system according to claim 1, wherein the recommendation generation means includes means for generating recommendations tailored to the user's situation using a trained model.

[1847] (Claim 6)

[1848] The system according to claim 1, wherein the user feedback means includes means for obtaining feedback from a user and transmitting it to a server.

[1849] (Claim 7)

[1850] The system according to claim 1, wherein the model improvement means includes means for personalizing and improving the accuracy of the model using user feedback and new data.

[1851] "Example 1"

[1852] (Claim 1)

[1853] Data collection means,

[1854] Data preprocessing means,

[1855] Learning methods using deep learning models,

[1856] A recommendation generation means for generating recommendations,

[1857] A user feedback mechanism for receiving user feedback,

[1858] A system that includes model improvement methods for improving the accuracy of a model.

[1859] (Claim 2)

[1860] The system according to claim 1, wherein the data collection means includes means for acquiring the user's past behavioral data and decision-making history.

[1861] (Claim 3)

[1862] The system according to claim 1, wherein the data preprocessing means includes means for cleansing and normalizing the acquired data.

[1863] "Application Example 1"

[1864] (Claim 1)

[1865] Data collection means,

[1866] Data preprocessing means,

[1867] Learning methods,

[1868] Recommended generation method,

[1869] User feedback methods,

[1870] Model improvement methods,

[1871] A means of acquiring and analyzing data to optimize the operational efficiency of robots within a factory,

[1872] A means of recommending the optimal operating schedule and maintenance timing,

[1873] A system that includes means for displaying recommended operational schedules and maintenance timings.

[1874] (Claim 2)

[1875] The system according to claim 1, wherein the data collection means includes means for acquiring, in addition to the user's past behavioral data and decision history, robot operation data, maintenance records, or production plans within the factory.

[1876] (Claim 3)

[1877] The system according to claim 1, wherein the data preprocessing means includes means for cleansing and normalizing the acquired data, as well as converting it into a format suitable for analyzing the operational efficiency of robots in a factory.

[1878] "Example 2 of combining an emotion engine"

[1879] (Claim 1)

[1880] Data collection means,

[1881] Data preprocessing means,

[1882] Learning methods,

[1883] Recommended generation method,

[1884] User feedback methods,

[1885] Model improvement methods,

[1886] A system that includes means for acquiring emotional data.

[1887] (Claim 2)

[1888] The system according to claim 1, wherein the data collection means includes means for acquiring the user's past behavioral data and decision-making history.

[1889] (Claim 3)

[1890] The system according to claim 1, wherein the data preprocessing means includes means for cleansing and normalizing the acquired data.

[1891] "Application example 2 when combining with an emotional engine"

[1892] (Claim 1)

[1893] Data collection means,

[1894] Data preprocessing means,

[1895] Learning methods,

[1896] Recommended generation method,

[1897] User feedback methods,

[1898] Model improvement methods,

[1899] Means of recognizing emotions,

[1900] Real-time decision support means,

[1901] A system that includes this.

[1902] (Claim 2)

[1903] The system according to claim 1, wherein the data collection means includes means for acquiring the user's past behavioral data and decision-making history.

[1904] (Claim 3)

[1905] The system according to claim 1, wherein the data preprocessing means includes means for cleansing and normalizing the acquired data. [Explanation of Symbols]

[1906] 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

[Claim 1] A data collection method for acquiring users' past behavioral data and decision-making history, A data preprocessing means for cleansing and normalizing acquired data, A learning method that uses a deep learning model to learn the user's decision-making process, A recommendation generation means that generates recommendations tailored to the user's situation using a pre-trained model, A user feedback mechanism that obtains user feedback and sends it to the server, A system that includes model improvement methods for personalizing and improving the accuracy of models using user feedback and new data.

Images