System

Abstract

A system is provided.SOLUTION: A system comprising: means for inputting a question related to a task by a user; means for receiving the question and generating an initial proposed answer by a generative model; means for presenting the initial proposed answer to a teacher; means for receiving a correction from the teacher and feeding back the correction to the generative model; and means for generating a final answer and providing the final answer to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Developing and updating business manuals in companies requires a significant amount of time and effort, placing a heavy burden on employees. Therefore, there is a need for a method to streamline the creation and updating of business manuals and minimize the amount of manual work required. Furthermore, current methods make it difficult to ensure the accuracy and consistency of information, and it is also difficult to quickly reflect new information and changes. To solve these issues, an effective and efficient system is needed. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides a system including a means for a user to input a business-related question, a means for receiving the question and generating an initial answer proposal using a generative model, a means for presenting the initial answer proposal to a teacher, a means for receiving corrections from the teacher and feeding them back to the generative model, and a means for generating a final answer and providing it to the user. Furthermore, the accuracy of the generative model can be improved by automatically updating and learning the generative model based on the feedback. Furthermore, by including a means for storing the final answer in a database and updating the generative model learning data and answer database through daily batch processing, the currency and accuracy of the information can be ensured. In this way, the present invention makes it possible to streamline the creation and updating of business manuals and minimize human labor.

[0006] A "user" is an individual or organization whose role is to utilize the system to enter business questions and receive generated answers.

[0007] The "teacher" is an individual or organization that reviews the generated initial answer proposals and provides corrections and feedback as necessary.

[0008] A "generative model" or "generative AI" is an artificial intelligence system that generates initial answers to questions received from users and automatically learns and updates based on feedback.

[0009] An "initial answer proposal" is the answer that a generative model first generates in response to a user's question.

[0010] "Feedback" refers to the teacher's correction of the initial answer or the provision of supplementary information.

[0011] The "final answer" is the final answer generated after the generative model is updated to reflect the teacher's feedback.

[0012] A "database" is a structured collection of data within a system that stores data such as user questions, generated answers, and teacher feedback.

[0013] "Batch processing" is a process in which a series of processes are executed in a system at once to update data or train models, especially at night. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram illustrating a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0015] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0018] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0019] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0020] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0025] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0027] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0034] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0035] This invention is a system that streamlines the creation and updating of business manuals within a company, and is composed of three main components: a user, a teacher, and a generative model. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[0036] User registration and role assignment

[0037] server:

[0038] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[0039] Device:

[0040] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[0041] User:

[0042] Enter the required information depending on your role (user or teacher).

[0043] Enter a question and generate initial answer ideas

[0044] User:

[0045] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[0046] Device:

[0047] The entered question is sent to the server, which receives it.

[0048] server:

[0049] The received question is passed to a generative model to generate an initial answer, which is then sent back to the user.

[0050] Teacher correction and feedback

[0051] User:

[0052] The user checks the initial answer proposal received from the server on the terminal and accesses an interface to present it to the teacher.

[0053] Device:

[0054] Provide a method for transferring initial answer ideas to the teacher.

[0055] server:

[0056] The corrections entered by the teacher are fed back to the generative model, and the model is updated.

[0057] Teacher:

[0058] The teacher reviews the initial answer and inputs any necessary corrections, which are then fed back to the generative model via the server.

[0059] Providing a final answer

[0060] server:

[0061] Based on the feedback, the generative model generates a final answer and provides it to the user.

[0062] Device:

[0063] The final answer received from the server is displayed to the user.

[0064] User:

[0065] The final answer will be confirmed and used as part of the business manual.

[0066] Nightly updates of information

[0067] server:

[0068] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[0069] Specific examples of operation

[0070] 1. User inquiry process:

[0071] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[0072] 2. Teacher feedback:

[0073] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[0074] 3. Final answer provided:

[0075] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[0076] In this way, the system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[0077] The processing flow will be explained below.

[0078] Program processing steps

[0079] Step 1: User registration and role assignment

[0080] server:

[0081] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[0082] Device:

[0083] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[0084] User:

[0085] Enter the required information depending on your role (user or teacher).

[0086] Step 2: Enter your question

[0087] User:

[0088] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[0089] Device:

[0090] The entered question is sent to the server.

[0091] server:

[0092] Receives a question and passes it to a generative model.

[0093] Step 3: Generate initial answer proposals

[0094] server:

[0095] Based on the received question, the generative model generates an initial answer proposal, which is returned to the user.

[0096] Device:

[0097] The initial answer plan received from the server is displayed to the user.

[0098] User:

[0099] Review the initial proposed answers.

[0100] Step 4: Present to the teacher

[0101] User:

[0102] A request for revision is sent to present the initial answer to the teacher.

[0103] Device:

[0104] Forward the correction request and initial answer proposal to the teacher.

[0105] server:

[0106] Relays correction requests and initial response proposals to be forwarded to the teacher.

[0107] Step 5: Teacher feedback

[0108] Teacher:

[0109] Check the initial response plan, enter any necessary corrections, and send the corrections to the server.

[0110] Device:

[0111] Enter the corrections made by the teacher and press the send button.

[0112] server:

[0113] The received corrections are fed back to the generative model to update the model.

[0114] Step 6: Generate and provide the final answer

[0115] server:

[0116] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[0117] Device:

[0118] The final answer received from the server is displayed to the user.

[0119] User:

[0120] The final answer will be confirmed and used as content for the business manual.

[0121] Step 7: Nightly update of information

[0122] server:

[0123] Batch processing is performed overnight to update the generative model training data and answer database, thereby maintaining the accuracy and consistency of the entire system.

[0124] Specific operation example

[0125] 1. User inquiry process:

[0126] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[0127] 2. Teacher feedback:

[0128] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[0129] 3. Final answer provided:

[0130] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[0131] As a result, this system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[0132] Example 1

[0133] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0134] Creating and updating business manuals within a company is time-consuming and labor-intensive, requiring a great deal of effort. Delays in updating information can also lead to problems with reduced business efficiency. In particular, there are limitations to the current manual process, which requires quick responses to questions while maintaining the quality of the answers. The purpose of this invention is to solve these problems and improve the efficiency of creating and updating business manuals.

[0135] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0136] In this invention, the server includes: means for a user to input a question related to a business; means for receiving the question and generating an initial answer using a generative AI model; means for the user to confirm the initial answer and present it to a teacher via a terminal; means for receiving corrections from the teacher and feeding them back to the generative AI model; means for generating a final answer and providing it to the user; and means for performing batch processing overnight to update the learning data and answer database of the generative AI model. This semi-automates the creation and updating of business manuals, significantly reducing the human labor required by companies while maintaining the accuracy and consistency of information.

[0137] A "user" is a person who operates the system and inputs questions about business.

[0138] A "terminal" is a device, such as a computer or mobile device, that a user uses to access the system.

[0139] The "server" is a central computer system that runs in a cloud environment and processes and manages questions from users and feedback from teachers.

[0140] A "generative AI model" is an artificial intelligence model that automatically generates answers to questions from users, and learns and updates based on feedback.

[0141] "Initial answer proposal" refers to the answer that the generative AI model first generates in response to a user's question.

[0142] The "teacher" is a person whose role is to review the initial proposed answers, input any necessary corrections, and provide feedback to the generative AI model.

[0143] "Final answer" refers to the answer that the generative AI model ultimately generates based on the teacher's feedback.

[0144] "Batch processing" refers to data updates and processing that are carried out in the background all at once during a certain period of time, such as at night.

[0145] A "business manual" refers to a document that describes the procedures, methods, rules, etc. for carrying out business.

[0146] A "database" is a data recording device or system for centrally managing and storing information within a system.

[0147] This invention is a system for streamlining the creation and updating of business manuals within a company. This system mainly consists of four elements: a server, a terminal, a user, and a generative AI model.

[0148] User registration and role assignment

[0149] server:

[0150] When the system is started for the first time, a user registration screen is displayed. This screen provides an interface where the administrator can designate the roles of "user" and "teacher." For example, when the administrator accesses "http: / / example.com / register," the user registration screen is displayed. The user information entered through this interface is saved in a database on the server.

[0151] Device:

[0152] The user accesses the registration screen and enters the required information (e.g., name, email address, role). For example, if the user enters the name "Ichiro Tanaka", email address "tanaka@example.com", and role "User", and clicks the send button, the information is sent to the server.

[0153] Enter a question and generate initial answer ideas

[0154] User:

[0155] Specific questions about the work are entered into the terminal interface. These questions are used to help create a work manual. For example, "How should I write a project progress report?"

[0156] Device:

[0157] The entered question is sent to the server.

[0158] server:

[0159] The received question is passed to a generative AI model, which generates an initial answer. The generated initial answer is then sent back to the user from the server. For example, if a generative AI model (e.g., GPT-4) is given the prompt, "How do I write a project progress report?", it will generate an initial answer saying, "A project progress report should include the following elements: progress, challenges, and next steps."

[0160] Teacher correction and feedback

[0161] User:

[0162] The terminal confirms the initial answer proposal received from the server and accesses an interface to present it to the teacher. For example, the user receives the answer "It is desirable to describe progress, challenges, and next steps," and transitions to a screen to send it to the teacher.

[0163] Device:

[0164] A notification will be sent to forward the initial answer to the teacher.

[0165] Teacher:

[0166] The teacher reviews the initial answer and inputs any necessary corrections. These corrections are fed back to the generative AI model via the server. For example, the teacher might correct the answer to "specifically, include the date, destination, purpose, outcome, and next action," and this is fed back to the generative AI model.

[0167] Providing a final answer

[0168] server:

[0169] Based on the teacher's feedback, the generative AI model generates a final answer and provides it to the user. For example, based on the updated feedback, the generative AI model generates a final answer such as "It is desirable for project progress reports to include dates, visits, objectives, results, and next actions."

[0170] Device:

[0171] The final answer is received from the server and displayed to the user. For example, the server may say, "It is desirable that the project progress report include the date, visit destination, objectives, results, and next actions," and the answer is displayed to the user.

[0172] Nightly updates of information

[0173] server:

[0174] Batch processing is performed at certain times during the night to update the generative AI model's learning data and answer database. This ensures the accuracy and consistency of the system. For example, the server starts batch processing every night at 2:00, updating the generative AI model's learning data with the latest feedback information.

[0175] This system streamlines the creation and updating of business manuals, reducing the amount of human effort required by companies while maintaining the accuracy and consistency of information.

[0176] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0177] Step 1:

[0178] server:

[0179] When the system is started for the first time, an interface is provided that displays a user registration screen. The administrator registers a new user and designates the role of "user" or "teacher." This registration information is stored in a database. For example, the administrator accesses "http: / / example.com / register," the user registration screen is displayed, and the entered information is sent to the server.

[0180] Input: User information entered by the administrator on the registration screen (e.g., name, email address, role)

[0181] Output: The registered user information is saved in the database.

[0182] Step 2:

[0183] Device:

[0184] The user accesses the registration screen through the device's browser and enters the required information (name, email address, role), which is then sent to the server.

[0185] Input: Information entered by the user (e.g., name "Ichiro Tanaka", email "tanaka@example.com", role "User")

[0186] Output: The input information is sent to the server.

[0187] Step 3:

[0188] server:

[0189] The received user information is stored in a database and the user's role is determined.

[0190] Input: User information sent from the device

[0191] Output: User information is saved in the database and roles are determined.

[0192] Step 4:

[0193] User:

[0194] Specific questions about the business are entered into the terminal interface and sent to the server.

[0195] Input: Question (e.g., "How do I write a project status report?")

[0196] Output: The entered question is sent to the server.

[0197] Step 5:

[0198] Device:

[0199] The entered question is sent to the server.

[0200] Input: The question entered by the user

[0201] Output: The question is sent to the server.

[0202] Step 6:

[0203] server:

[0204] The received question is passed to the generative AI model to generate an initial answer, which is then sent back to the user from the server.

[0205] Input: User-submitted question

[0206] Output: An initial answer proposed by the generative AI model (e.g., "Project status reports should include the following elements: progress, challenges, and next steps.")

[0207] Step 7:

[0208] User:

[0209] The initial answer plan received from the server is checked on the terminal and an interface is accessed to present it to the teacher.

[0210] Input: Initial answer

[0211] Output: Display the initial answer plan in the interface to present it to the teacher.

[0212] Step 8:

[0213] Device:

[0214] Send a notification to forward the initial answer to the teacher.

[0215] Input: Initial answer

[0216] Output: Notification to forward to teacher

[0217] Step 9:

[0218] Teacher:

[0219] The user checks the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative AI model via the server.

[0220] Input: Initial answer, revisions (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action")

[0221] Output: The corrections are fed back into the generative AI model

[0222] Step 10:

[0223] server:

[0224] Based on the teacher's feedback, the generative AI model generates the final answer and provides it to the user.

[0225] Input: Teacher feedback

[0226] Output: Final proposed answer (e.g., "Project progress reports should include dates, locations, objectives, outcomes, and next actions.")

[0227] Step 11:

[0228] Device:

[0229] The final answer is received from the server and displayed to the user.

[0230] Input: Final answer

[0231] Output: Display the final proposed answer to the user

[0232] Step 12:

[0233] User:

[0234] The final answer will be confirmed and used as part of the business manual.

[0235] Input: Final answer

[0236] Output: Final answer reflected in the business manual

[0237] Step 13:

[0238] server:

[0239] Nightly batch processing is performed to update the training data and response database for the generative AI model, ensuring the accuracy and consistency of the system.

[0240] Input: Feedback information, response database

[0241] Output: Updated generative AI model and database

[0242] In this way, the system concretely performs each processing step and efficiently supports the creation and updating of business manuals.

[0243] (Application example 1)

[0244] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0245] Creating and updating operational manuals within a logistics center requires a lot of time and manpower, resulting in low work efficiency. Furthermore, as the content of operations changes daily, it is difficult to keep up with the latest information, resulting in a decline in work efficiency and an increased risk of mistakes. Therefore, a system is needed to semi-automate the creation and updating of operational manuals, thereby reducing human labor and improving work efficiency.

[0246] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0247] In this invention, the server includes means for a user to input questions related to business operations, means for receiving the questions and generating initial answer proposals using a generative model, means for presenting the initial answer proposals to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating final answers and providing them to the user, means for automatically updating the generative model and database overnight, and means for streamlining the input and generation of questions and answers specialized for logistics operations. This enables the efficient creation and updating of business manuals within a logistics center, improving work efficiency and reducing human labor.

[0248] A "user" is a person who uses the system to input questions about business and operates a terminal that receives answers.

[0249] The "teacher" is a person whose role is to check the initial answer plan and input any necessary corrections.

[0250] A "generative model" is an AI algorithm that generates initial answer proposals based on input questions and then learns and updates them based on feedback.

[0251] A "question" is an inquiry entered by a user about information or procedures related to a business.

[0252] An "initial answer proposal" is a proposal of an answer that a generative model first generates based on a question from a user.

[0253] "Correction" refers to the teacher's correction or addition to the initial answer plan.

[0254] "Feedback" is the process in which the teacher communicates the corrections to the generative model and incorporates them as training data for the model.

[0255] The "final answer" is the answer that the generative model ultimately provides after reflecting the corrections and feedback.

[0256] "Nightly update" is a process that automatically updates the generative model and database at night.

[0257] "Logistics operations" refers to all operations related to the handling, storage, shipping, etc. of goods carried out within a logistics center.

[0258] This invention is a system that streamlines the creation and updating of business manuals within a company, and is particularly applicable to operations at logistics centers. The system has three main components: a user, a teacher, and a generative model, and runs on a cloud server. Users and teachers can access the system via their smartphones.

[0259] User registration and role assignment

[0260] The server displays a user registration screen when the system is started for the first time, and provides an interface for the administrator to specify the roles of "user" and "teacher." The role information of the designated user is stored in the database.

[0261] The terminal provides a means for the user to access the registration screen and input the necessary information, which is then sent to the server.

[0262] The user enters the necessary information according to his / her role (user or teacher).

[0263] Enter a question and generate initial answer ideas

[0264] Users input questions about logistics operations into the terminal interface, and these questions are used to create an operations manual.

[0265] The terminal sends the entered question to the server, which passes it to the generative model.

[0266] The server passes the received question to the generative model to generate an initial answer, which is then sent back to the user via the device.

[0267] Teacher correction and feedback

[0268] The user checks the initial answer plan received from the server on the terminal and accesses an interface for presenting it to the teacher.

[0269] The terminal provides a means for transferring the initial answer proposal to the teacher.

[0270] The server feeds back the corrections entered by the teacher to the generative model and updates the model.

[0271] The teacher reviews the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative model via the server.

[0272] Providing a final answer

[0273] The server uses the generative model to generate a final answer based on the feedback and provides it to the user via the terminal.

[0274] The terminal displays the final answer received from the server to the user.

[0275] The user checks the final answer and uses it as content for the business manual.

[0276] Nightly updates of information

[0277] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[0278] Specific examples of operation

[0279] User Question Process:

[0280] When a user inputs a question such as "What can we do to improve the efficiency of logistics operations?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "One way to improve efficiency is to increase the speed at which products are picked."

[0281] Teacher feedback:

[0282] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, zoning the work area, using RFID tags, and providing regular training would be effective."). The server then feeds these modifications back into the generative model, updating it.

[0283] Final answer provided:

[0284] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in their work manual, saying, "To improve the efficiency of logistics operations, it is effective to zone the work area, use RFID tags, and conduct regular training."

[0285] This will enable the efficient creation and updating of business manuals within the logistics center, improving work efficiency and reducing human labor.

[0286] Example prompt sentence:

[0287] How can we improve the efficiency of picking operations at logistics centers?

[0288] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0289] Step 1:

[0290] Users input business-related questions through a smartphone application.

[0291] input:

[0292] Questions (e.g., "What can we do to improve the efficiency of our logistics operations?")

[0293] output:

[0294] Entered question data

[0295] Step 2:

[0296] The terminal sends the entered question to the server.

[0297] input:

[0298] Question data entered by the user

[0299] output:

[0300] Question data sent to the server

[0301] Step 3:

[0302] The server receives the question and passes it to the generative AI model.

[0303] input:

[0304] Question data submitted by users

[0305] output:

[0306] Question data passed to the generative AI model

[0307] Step 4:

[0308] A generative AI model generates initial answer suggestions based on the question.

[0309] input:

[0310] Question Data

[0311] Data processing:

[0312] Infer the best answer to a question using the training data inside the model

[0313] output:

[0314] Initial answer (e.g., "One way to improve efficiency is to increase the speed at which products are picked.")

[0315] Step 5:

[0316] The server receives the generated initial answer plan and transmits it to the user via the terminal.

[0317] input:

[0318] Initial draft answer

[0319] output:

[0320] The initial answer plan data is sent to the user's device.

[0321] Step 6:

[0322] The user reviews the initial answer plan and presents it to the teacher.

[0323] input:

[0324] Initial draft answer

[0325] output:

[0326] Data to present to the teacher

[0327] Step 7:

[0328] The teacher inputs corrections to the initial answer.

[0329] input:

[0330] Initial response proposal and revisions (e.g., "Specifically, zoning work areas, using RFID tags, and providing regular training are effective.")

[0331] output:

[0332] Correction data

[0333] Step 8:

[0334] The device sends the corrections made by the teacher to the server.

[0335] input:

[0336] Correction data from the teacher

[0337] output:

[0338] Modification data sent to the server

[0339] Step 9:

[0340] The server receives the corrections and feeds them back into the generative model.

[0341] input:

[0342] Correction data

[0343] output:

[0344] Data fed back into the generative model

[0345] Step 10:

[0346] The generative model learns and updates based on feedback.

[0347] input:

[0348] Feedback correction data

[0349] Data Calculation:

[0350] The process of incorporating the corrections into the training data and updating the model

[0351] output:

[0352] Updated generative model

[0353] Step 11:

[0354] The server uses the updated generative model to generate a final answer and provides it to the user.

[0355] input:

[0356] Updated generative model

[0357] Data processing:

[0358] The process of reflecting newly learned content and generating the final answer

[0359] output:

[0360] Final response data (e.g., "Zoning work areas, using RFID tags, and providing regular training are effective ways to improve the efficiency of logistics operations.")

[0361] Step 12:

[0362] The terminal displays the final answer received from the server to the user.

[0363] input:

[0364] Final response data

[0365] output:

[0366] Final response data displayed on the user's device

[0367] Step 13:

[0368] The server performs batch processing overnight to update the generative model's training data and response database to the latest information.

[0369] input:

[0370] All feedback data collected during the day

[0371] Data Calculation:

[0372] Batch processing to capture all feedback data and update the model and database

[0373] output:

[0374] Updated generative models and answer database

[0375] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0376] This invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[0377] User registration and role assignment

[0378] server:

[0379] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[0380] Device:

[0381] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[0382] User:

[0383] Enter the required information depending on your role (user or teacher).

[0384] Question input and emotion recognition

[0385] User:

[0386] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[0387] Device:

[0388] The entered question is sent to the server, and the emotion engine analyzes the user's emotional state.

[0389] server:

[0390] The system receives a question and obtains the user's emotional information analyzed by the emotion engine.Then, the question and emotional information are passed to the generative model to generate an initial answer.

[0391] Generate initial answer proposals

[0392] server:

[0393] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[0394] Device:

[0395] The initial answer plan received from the server is displayed to the user.

[0396] User:

[0397] Review the initial proposed answers.

[0398] Presentation to the teacher and provision of emotional information

[0399] User:

[0400] A request for revision is sent to present the initial answer to the teacher.

[0401] Device:

[0402] The correction request, the initial response plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[0403] server:

[0404] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[0405] Teacher feedback

[0406] Teacher:

[0407] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[0408] Device:

[0409] Enter the corrections made by the teacher and press the send button.

[0410] server:

[0411] The received corrections are fed back to the generative model to update the model.

[0412] Generate and provide the final answer

[0413] server:

[0414] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[0415] Device:

[0416] The final answer received from the server is displayed to the user.

[0417] User:

[0418] The final answer will be confirmed and used as content for the business manual.

[0419] Nightly updates of information

[0420] server:

[0421] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[0422] Specific operation example

[0423] 1. User inquiry process:

[0424] When a user inputs a question such as "How should a sales report be written?", the server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to a generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[0425] 2. Teacher feedback:

[0426] The user receives the initial answer and presents it to the teacher, who then inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotion information. The server then feeds these corrections back to the generative model, updating it.

[0427] 3. Final answer provided:

[0428] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[0429] In this way, the system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[0430] The processing flow will be explained below.

[0431] Program processing steps

[0432] Step 1: User registration and role assignment

[0433] server:

[0434] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[0435] Device:

[0436] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[0437] User:

[0438] Enter the required information depending on your role (user or teacher).

[0439] Step 2: Enter your question

[0440] User:

[0441] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[0442] Device:

[0443] The entered question is sent to the server.

[0444] server:

[0445] Receives a question and passes the question content to the emotion engine.

[0446] Step 3: Emotion Recognition

[0447] server:

[0448] The emotion engine analyzes the emotional state from the user's input and obtains the results. The emotion information and question are passed to the generative model.

[0449] Step 4: Generate initial answer proposals

[0450] server:

[0451] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[0452] Device:

[0453] The initial answer plan received from the server is displayed to the user.

[0454] User:

[0455] Review the initial proposed answers.

[0456] Step 5: Presenting to the teacher and providing emotional information

[0457] User:

[0458] A request for revision is sent to present the initial answer to the teacher.

[0459] Device:

[0460] The correction request, the initial answer plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[0461] server:

[0462] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[0463] Step 6: Teacher feedback

[0464] Teacher:

[0465] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[0466] Device:

[0467] Enter the corrections made by the teacher and press the send button.

[0468] server:

[0469] The received corrections are fed back to the generative model to update the model.

[0470] Step 7: Generate and deliver the final answer

[0471] server:

[0472] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[0473] Device:

[0474] The final answer received from the server is displayed to the user.

[0475] User:

[0476] The final answer will be confirmed and used as content for the business manual.

[0477] Step 8: Update information overnight

[0478] server:

[0479] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[0480] Specific operation example

[0481] 1. User inquiry process:

[0482] When a user inputs a question such as "How do I write a sales report?", the device sends it to the server. The server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to the generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable for a sales report to include the following elements."

[0483] 2. Teacher feedback:

[0484] After receiving the initial answer, the user submits a correction request to present it to the teacher. The server forwards the initial answer along with the user's emotional state to the teacher. The teacher inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotional information. The server feeds these corrections back into the generative model, updating it.

[0485] 3. Final answer provided:

[0486] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[0487] As described above, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[0488] Example 2

[0489] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0490] The creation and updating of business manuals has traditionally been a time-consuming and labor-intensive process, requiring a great deal of manual work. It has also been difficult to maintain consistency and accuracy in answers to questions, and there have been cases where the right answer was not given to a particular question. To solve this problem, a system is needed that can quickly and accurately update business manuals while taking into account the emotional state of the user.

[0491] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0492] In this invention, the server includes means for displaying a user registration screen when the system is first started and for users to register roles including an administrator role, means for users to input questions related to their work, means for receiving the questions and analyzing the user's emotional state with an emotion engine, and means for generating initial answer proposals with a generative model based on the questions and the analyzed emotion information. This makes it possible to provide quick and accurate answers to questions that take the user's emotional state into consideration.

[0493] A "user" is a general user of this system, who inputs questions about business.

[0494] An "administrator" is a person who has the authority to access the system and to designate users and teachers through the user registration screen when the system is started for the first time.

[0495] "Questions about business" are questions about content related to the creation and updating of business manuals within a company, and are data that are entered into the system.

[0496] An "emotion engine" is a software component that analyzes a user's emotional state in response to a question entered by the user.

[0497] A "generative model" is a machine learning algorithm that generates initial and final answers based on the input question and emotional information.

[0498] An "initial answer proposal" is an answer that is initially generated by the generative model based on the question and emotion information from the user.

[0499] The "teacher" is a person whose role is to receive the initial answer proposal from the user and make any necessary corrections.

[0500] A "request for correction" is a request that a user submits to the teacher to request correction of the initial answer plan.

[0501] The "final answer" is the answer that the generative model finally generates based on the corrections made by the teacher.

[0502] "Batch processing" is a series of processes that are executed overnight and are used to update the training data and answer database for the generative model.

[0503] The present invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[0504] Hardware and software used

[0505] This system is composed of the following hardware and software.

[0506] Server: A cloud server is used to display the user registration screen, receive questions, analyze emotions, generate suggested answers, relay correction requests, and provide final answers.

[0507] Terminal: A device used by a user, such as a computer or smartphone, that provides an interface for the user to enter questions or correction requests.

[0508] Generative model: A model that uses machine learning algorithms to generate initial and final answers based on the input question and sentiment information.

[0509] Emotion Engine: A software component that uses natural language processing techniques to parse the emotional state from a user's question.

[0510] System operation procedure

[0511] The operating procedure of this system is as follows.

[0512] 1. User Registration:

[0513] When the system is started for the first time, the server displays a user registration screen. The administrator accesses the screen via a terminal and specifies the user information and role (administrator, user, teacher). This information is saved in the database.

[0514] 2. Question input and emotion recognition:

[0515] The user inputs a business-related question into the device interface. The device then sends the question to the server, which then uses an emotion engine to analyze the user's emotional state. The analyzed emotion information is then provided to the generative model.

[0516] 3. Generate initial answer:

[0517] The server passes the question and the analyzed emotion information to the generative model to generate initial answer proposals, which are then displayed on the user's device.

[0518] 4. Request for correction:

[0519] The user checks the initial answer plan, and if they are dissatisfied, they send a correction request to the teacher. This request also includes the user's emotional information. The device sends the correction request and emotional information to the server, and the server forwards them to the teacher.

[0520] 5. Teacher feedback:

[0521] The teacher reviews the initial answer and emotion information, inputs any necessary corrections, and sends them to the server. The server then feeds the received corrections back to the generative model, which then generates the final answer.

[0522] 6. Providing a Final Response:

[0523] The generated final answer is sent from the server to the user terminal, and the user reflects the final answer in the business manual.

[0524] 7. Nightly updates:

[0525] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[0526] Specific operation example

[0527] A specific example of operation is shown below.

[0528] 1. User inquiry process:

[0529] The user inputs a question such as "How do I write a sales report?" The server uses an emotion engine to analyze the user's emotional state. The question along with the analysis results is sent to the generative model, which generates an initial answer: "A sales report should include the following elements: dates, destinations, objectives, results, and next actions."

[0530] 2. Teacher feedback:

[0531] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[0532] 3. Final answer provided:

[0533] Based on the revised content, the generative model generates the final answer: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Specific examples should also be added, such as 'Date: 2023 / 10 / 01'." The server then provides the final answer to the user.

[0534] In this way, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[0535] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0536] Step 1:

[0537] User registration and role assignment

[0538] Server: Display the user registration screen when the system is started for the first time.

[0539] Input: System first boot.

[0540] Specific operation: The server generates a registration screen that can be accessed on a browser and displays a "New User Registration" button.

[0541] Output: Display of registration screen.

[0542] Terminal: The administrator accesses the registration screen and enters user information and roles (user role, teacher role).

[0543] Input: Enter your user information (name, email address, password) and role.

[0544] Specific operation: The administrator enters information such as "Yamada Taro," "taro@example.com," and "Teacher role."

[0545] Output: Sending user information and roles.

[0546] Server: Receives input information and stores role information in a database.

[0547] Input: User information and role.

[0548] Specific operation: The server stores information such as "Yamada Taro," "taro@example.com," and "teacher" in a database.

[0549] Output: User information and roles are saved in the database.

[0550] Step 2:

[0551] Question input and emotion recognition

[0552] User: Enters business-related questions into the terminal interface.

[0553] Input: The question.

[0554] Specific Action: User types "How do I write a sales report?"

[0555] Output: Question data.

[0556] Terminal: Sends the entered question to the server.

[0557] Input: Question data.

[0558] Specific operation: The terminal sends the question data "How do you write a report for sales activities?" to the server.

[0559] Output: Sends the query data to the server.

[0560] Server: Analyzes the user's emotional state using the emotion engine.

[0561] Input: Question data.

[0562] Specific operation: The emotion engine determines the user's stress level as "medium" from the question.

[0563] Output: Emotional information.

[0564] Step 3:

[0565] Generate initial answer proposals

[0566] Server: Requests a generative model based on the question and emotion information to generate an initial answer.

[0567] Input: Question data and sentiment information.

[0568] Specific behavior: The generative AI model generates the following: "Sales activity reports should include the following elements: date, visit, purpose, results, and next actions."

[0569] Output: Initial draft answer.

[0570] Terminal: Show initial answer ideas to the user.

[0571] Input: Initial answer proposal.

[0572] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, destinations, objectives, results, and next actions."

[0573] Output: Display of initial answer proposal to user.

[0574] Step 4:

[0575] Submitting a revision request

[0576] User: If dissatisfied with the initial answer, send a request for revision to the teacher.

[0577] Input: Initial response proposal, requested revisions.

[0578] Specific action: The user sends a correction request with the comment "I would like more specific examples."

[0579] Output: Submit a correction request.

[0580] Terminal: Sends the initial answer proposal, including the request for revision and emotional information, to the teacher.

[0581] Input: Request for revision, initial response plan, sentiment information.

[0582] Specific action: The device sends the comment "I want more concrete examples" and the user's emotional information to the server, which then forwards it to the teacher.

[0583] Output: Transfer to teacher.

[0584] Step 5:

[0585] Teacher feedback

[0586] Teacher: Check the initial answer and emotional information provided and enter any corrections.

[0587] Input: Initial answer proposal, sentiment information, revisions.

[0588] Specific Action: The teacher types, "In your sales activity report, include the date, destination, objective, results, and next action. Also, add specific examples to each section."

[0589] Output: The corrections.

[0590] On your device: Click the Modify button to send the modifications to the server.

[0591] Input: The correction.

[0592] Specific actions: The teacher enters the corrections and clicks the submit button.

[0593] Output: Send to server.

[0594] Server: Feedback the received corrections to the generative model and update the model.

[0595] Input: The correction.

[0596] Specific operation: The server feeds the corrections back to the generative model, and the model is updated.

[0597] Output: An updated generative model.

[0598] Step 6:

[0599] Generate and provide the final answer

[0600] Server: Based on the teacher's feedback, the generative model generates the final answer.

[0601] Input: The updated generative model.

[0602] Specific behavior: The generative AI model generates, "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[0603] Output: The final answer.

[0604] Terminal: Display the final answer to the user.

[0605] Input: Final answer.

[0606] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[0607] Output: What is displayed to the user.

[0608] Step 7:

[0609] Nightly updates of information

[0610] Server: Performs batch processing overnight to update the generative model training data and answer database to the latest state.

[0611] Input: Latest fix information.

[0612] Specific operation: Batch processing is performed automatically at 11:00 PM, and the generative model adds new correction information to the training data.

[0613] Output: Updated training data and answer database.

[0614] (Application example 2)

[0615] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0616] Creating and updating manuals for factory operations requires a great deal of effort and time. Operations procedures for industrial machinery, in particular, involve many steps, and require frequent manual correction and verification, resulting in reduced efficiency and increased error rates. It is also important to ensure that employees accurately understand and follow the appropriate procedures. Failure to do so can lead to reduced productivity and potential quality issues.

[0617] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0618] In this invention, the server includes means for a user to input a question related to a business, means for receiving the question and generating an initial answer proposal using a generative model, means for presenting the initial answer proposal to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating a final answer and providing it to the user, means for providing analyzed emotion information to the generative model, and means for generating work procedures for industrial machinery and providing the work procedures to the user and the industrial machinery. This enables efficient and accurate creation and updating of work procedure manuals.

[0619] A "user" is a person who utilizes the system to enter business-related questions and review the generated answers.

[0620] The "teacher" is a person whose role is to review the initial answer proposals generated by the generative model and provide any necessary corrections or feedback.

[0621] A "generative model" is an artificial intelligence model that generates initial answers to input questions and automatically learns and updates based on feedback.

[0622] "Emotion information" is data related to emotions analyzed from questions entered by users, and is information that influences the generation of answers by the generative model.

[0623] "Industrial machinery" refers to production equipment and work robots used in factories.

[0624] "Work procedures" are guidelines that describe specific steps for carrying out a particular task or operation.

[0625] The "server" is a central computer that manages and operates the entire system, receiving questions, generating answer proposals, analyzing emotional information, and processing feedback.

[0626] The "database" is an information management system for storing and managing the generated final answers and feedback information.

[0627] "Batch processing" is a method for automatically updating and processing database information at regular intervals.

[0628] This invention is a system that enables factory robots and workers to efficiently create and update work procedures. The main components include a server, terminals, users, a teacher, a generative model, an emotion engine, and industrial machinery. This system operates as follows:

[0629] Server Features

[0630] The server is the central control and performs the following main tasks:

[0631] 1. Question reception and emotion recognition: The system receives a question entered by the user and analyzes the emotion information related to the question using an emotion engine. An emotion engine such as IBM Watson is sometimes used.

[0632] 2. Initial answer generation: The question along with the emotion information is passed to a generative model to generate an initial answer. OpenAI GPT-3 and other models are sometimes used as generative AI models.

[0633] 3. Corrective feedback: The corrective feedback from the teacher is received and fed back to the generative model. This feedback allows the generative model to learn and update automatically.

[0634] 4. Final answer generation: A final answer is generated based on the feedback and provided to the user.

[0635] 5. Information update: The final answers and related data are saved in the database and processed in daily batches.

[0636] Device Role

[0637] The terminal acts as a link between the user and the server and performs the following tasks:

[0638] 1. Providing a question input interface: Provide an interface that allows users to input questions about their work.

[0639] 2. Display answers: The initial answer proposals and final answers received from the server are displayed to the user.

[0640] 3. Sending correction request to teacher: Sends the user's correction request to the server.

[0641] User Roles

[0642] Users interact with the system as follows:

[0643] 1. Question input: A question about the business is input into the terminal and sent to the server.

[0644] 2. Confirmation of answers: Confirm the initial and final answers and request corrections if necessary.

[0645] 3. Use of the final answer: Use the final answer as an operations manual.

[0646] The role of the teacher

[0647] The teacher reviews the initial answer proposal and emotion information and provides any necessary corrections to the server, which then updates the generative model based on this information.

[0648] Usage example

[0649] For example, a user might input a question such as, "Please tell me the assembly procedure for a new product." The server uses an emotion engine to analyze the user's emotional state and passes this information along with the question to the generative model. The generative model generates an initial answer proposal, "The assembly procedure for a new product is as follows...," and provides this to the user. The user presents this initial answer proposal to the teacher, who then provides a correction such as, "Specifically, please also include the amount of force used when tightening the screws." The server feeds this correction back into the generative model, which then generates a final answer. The final answer is, "The assembly procedure for a new product is as follows...Please also include the amount of force used when tightening the screws."

[0650] Prompt Sentence Examples

[0651] Input: "What are the assembly steps for my new product?"

[0652] Initial response: "Here are the assembly steps for your new product..."

[0653] Feedback: "Please also include details on how much force to use when tightening the screws."

[0654] In this way, the system can provide efficient and accurate production and update of work procedures in industrial settings.

[0655] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0656] Step 1:

[0657] The user inputs a question about the job into the terminal. The input data is a specific question related to the job. For example, the user might input, "Please tell me the assembly procedure for a new product." The terminal then sends this input data to the server.

[0658] Step 2:

[0659] The server analyzes the question received from the user. First, it uses an emotion engine (e.g., IBM Watson) to analyze the emotional information contained in the question. Emotional information is data that indicates the user's level of urgency and importance regarding the question. This emotional information and the question content are passed to a generative AI model (e.g., OpenAI GPT-3).

[0660] Step 3:

[0661] The server generates an initial answer using a generative AI model. The generative AI model generates an appropriate answer based on the user's question and emotional information. For example, it generates an initial answer such as, "The assembly procedure for a new product is as follows..." This initial answer is returned from the server to the user via the device.

[0662] Step 4:

[0663] The user checks the initial answer provided by the server. After checking, the user inputs a correction request if necessary and sends it to the server via the terminal. The correction request includes specific corrections, such as "Specifically, please also include the amount of force used when tightening the screws."

[0664] Step 5:

[0665] The server forwards the correction request received from the user to the teacher, who then checks the correction request and the initial answer plan and enters any necessary corrections or supplementary information. The teacher's feedback is then sent back to the server.

[0666] Step 6:

[0667] The server receives corrective feedback from the teacher and sends it back to the generative AI model. This feedback updates the generative model, improving its accuracy. The generative AI model then generates a new answer based on the feedback and provides the final answer.

[0668] Step 7:

[0669] The final answer is transferred from the server to the terminal and provided to the user. For example, it might be, "The assembly procedure for a new product is as follows... Please also include the amount of force to use when tightening the screws." The user uses this final answer as their work manual.

[0670] Step 8:

[0671] At night, the server runs batch processing to update the information in the database. This keeps the generative model training data and the final answer database up to date. Updating the database is important for maintaining the accuracy and consistency of the entire system.

[0672] In this way, the server generates initial answer proposals using the user's question and emotional information, and then provides an accurate final answer after receiving feedback from a teacher.

[0673] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0674] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0675] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0676] [Second embodiment]

[0677] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0678] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0679] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0680] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0681] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0682] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0683] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0684] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0685] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0686] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0687] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0688] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0689] This invention is a system that streamlines the creation and updating of business manuals within a company, and is composed of three main components: a user, a teacher, and a generative model. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[0690] User registration and role assignment

[0691] server:

[0692] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[0693] Device:

[0694] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[0695] User:

[0696] Enter the required information depending on your role (user or teacher).

[0697] Enter a question and generate initial answer ideas

[0698] User:

[0699] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[0700] Device:

[0701] The entered question is sent to the server, which receives it.

[0702] server:

[0703] The received question is passed to a generative model to generate an initial answer, which is then sent back to the user.

[0704] Teacher correction and feedback

[0705] User:

[0706] The user checks the initial answer proposal received from the server on the terminal and accesses an interface to present it to the teacher.

[0707] Device:

[0708] Provide a method for transferring initial answer ideas to the teacher.

[0709] server:

[0710] The corrections entered by the teacher are fed back to the generative model, and the model is updated.

[0711] Teacher:

[0712] The teacher reviews the initial answer and inputs any necessary corrections, which are then fed back to the generative model via the server.

[0713] Providing a final answer

[0714] server:

[0715] Based on the feedback, the generative model generates a final answer and provides it to the user.

[0716] Device:

[0717] The final answer received from the server is displayed to the user.

[0718] User:

[0719] The final answer will be confirmed and used as part of the business manual.

[0720] Nightly updates of information

[0721] server:

[0722] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[0723] Specific examples of operation

[0724] 1. User inquiry process:

[0725] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[0726] 2. Teacher feedback:

[0727] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[0728] 3. Final answer provided:

[0729] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[0730] In this way, the system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[0731] The processing flow will be explained below.

[0732] Program processing steps

[0733] Step 1: User registration and role assignment

[0734] server:

[0735] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[0736] Device:

[0737] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[0738] User:

[0739] Enter the required information depending on your role (user or teacher).

[0740] Step 2: Enter your question

[0741] User:

[0742] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[0743] Device:

[0744] The entered question is sent to the server.

[0745] server:

[0746] Receives a question and passes it to a generative model.

[0747] Step 3: Generate initial answer proposals

[0748] server:

[0749] Based on the received question, the generative model generates an initial answer proposal, which is returned to the user.

[0750] Device:

[0751] The initial answer plan received from the server is displayed to the user.

[0752] User:

[0753] Review the initial proposed answers.

[0754] Step 4: Present to the teacher

[0755] User:

[0756] A request for revision is sent to present the initial answer to the teacher.

[0757] Device:

[0758] Forward the correction request and initial answer proposal to the teacher.

[0759] server:

[0760] Relays correction requests and initial response proposals to be forwarded to the teacher.

[0761] Step 5: Teacher feedback

[0762] Teacher:

[0763] Check the initial response plan, enter any necessary corrections, and send the corrections to the server.

[0764] Device:

[0765] Enter the corrections made by the teacher and press the send button.

[0766] server:

[0767] The received corrections are fed back to the generative model to update the model.

[0768] Step 6: Generate and provide the final answer

[0769] server:

[0770] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[0771] Device:

[0772] The final answer received from the server is displayed to the user.

[0773] User:

[0774] The final answer will be confirmed and used as content for the business manual.

[0775] Step 7: Nightly update of information

[0776] server:

[0777] Batch processing is performed overnight to update the generative model training data and answer database, thereby maintaining the accuracy and consistency of the entire system.

[0778] Specific operation example

[0779] 1. User inquiry process:

[0780] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[0781] 2. Teacher feedback:

[0782] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[0783] 3. Final answer provided:

[0784] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[0785] As a result, this system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[0786] Example 1

[0787] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0788] Creating and updating business manuals within a company is time-consuming and labor-intensive, requiring a great deal of effort. Delays in updating information can also lead to problems with reduced business efficiency. In particular, there are limitations to the current manual process, which requires quick responses to questions while maintaining the quality of the answers. The purpose of this invention is to solve these problems and improve the efficiency of creating and updating business manuals.

[0789] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0790] In this invention, the server includes: means for a user to input a question related to a business; means for receiving the question and generating an initial answer using a generative AI model; means for the user to confirm the initial answer and present it to a teacher via a terminal; means for receiving corrections from the teacher and feeding them back to the generative AI model; means for generating a final answer and providing it to the user; and means for performing batch processing overnight to update the learning data and answer database of the generative AI model. This semi-automates the creation and updating of business manuals, significantly reducing the human labor required by companies while maintaining the accuracy and consistency of information.

[0791] A "user" is a person who operates the system and inputs questions about business.

[0792] A "terminal" is a device, such as a computer or mobile device, that a user uses to access the system.

[0793] The "server" is a central computer system that runs in a cloud environment and processes and manages questions from users and feedback from teachers.

[0794] A "generative AI model" is an artificial intelligence model that automatically generates answers to questions from users, and learns and updates based on feedback.

[0795] "Initial answer proposal" refers to the answer that the generative AI model first generates in response to a user's question.

[0796] The "teacher" is a person whose role is to review the initial proposed answers, input any necessary corrections, and provide feedback to the generative AI model.

[0797] "Final answer" refers to the answer that the generative AI model ultimately generates based on the teacher's feedback.

[0798] "Batch processing" refers to data updates and processing that are carried out in the background all at once during a certain period of time, such as at night.

[0799] A "business manual" refers to a document that describes the procedures, methods, rules, etc. for carrying out business.

[0800] A "database" is a data recording device or system for centrally managing and storing information within a system.

[0801] This invention is a system for streamlining the creation and updating of business manuals within a company. This system mainly consists of four elements: a server, a terminal, a user, and a generative AI model.

[0802] User registration and role assignment

[0803] server:

[0804] When the system is started for the first time, a user registration screen is displayed. This screen provides an interface where the administrator can designate the roles of "user" and "teacher." For example, when the administrator accesses "http: / / example.com / register," the user registration screen is displayed. The user information entered through this interface is saved in a database on the server.

[0805] Device:

[0806] The user accesses the registration screen and enters the required information (e.g., name, email address, role). For example, if the user enters the name "Ichiro Tanaka", email address "tanaka@example.com", and role "User", and clicks the send button, the information is sent to the server.

[0807] Enter a question and generate initial answer ideas

[0808] User:

[0809] Specific questions about the work are entered into the terminal interface. These questions are used to help create a work manual. For example, "How should I write a project progress report?"

[0810] Device:

[0811] The entered question is sent to the server.

[0812] server:

[0813] The received question is passed to a generative AI model, which generates an initial answer. The generated initial answer is then sent back to the user from the server. For example, if a generative AI model (e.g., GPT-4) is given the prompt, "How do I write a project progress report?", it will generate an initial answer saying, "A project progress report should include the following elements: progress, challenges, and next steps."

[0814] Teacher correction and feedback

[0815] User:

[0816] The terminal confirms the initial answer proposal received from the server and accesses an interface to present it to the teacher. For example, the user receives the answer "It is desirable to describe progress, challenges, and next steps," and transitions to a screen to send it to the teacher.

[0817] Device:

[0818] A notification will be sent to forward the initial answer to the teacher.

[0819] Teacher:

[0820] The teacher reviews the initial answer and inputs any necessary corrections. These corrections are fed back to the generative AI model via the server. For example, the teacher might correct the answer to "specifically, include the date, destination, purpose, outcome, and next action," and this is fed back to the generative AI model.

[0821] Providing a final answer

[0822] server:

[0823] Based on the teacher's feedback, the generative AI model generates a final answer and provides it to the user. For example, based on the updated feedback, the generative AI model generates a final answer such as "It is desirable for project progress reports to include dates, visits, objectives, results, and next actions."

[0824] Device:

[0825] The final answer is received from the server and displayed to the user. For example, the server may say, "It is desirable that the project progress report include the date, visit destination, objectives, results, and next actions," and the answer is displayed to the user.

[0826] Nightly updates of information

[0827] server:

[0828] Batch processing is performed at certain times during the night to update the generative AI model's learning data and answer database. This ensures the accuracy and consistency of the system. For example, the server starts batch processing every night at 2:00, updating the generative AI model's learning data with the latest feedback information.

[0829] This system streamlines the creation and updating of business manuals, reducing the amount of human effort required by companies while maintaining the accuracy and consistency of information.

[0830] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0831] Step 1:

[0832] server:

[0833] When the system is started for the first time, an interface is provided that displays a user registration screen. The administrator registers a new user and designates the role of "user" or "teacher." This registration information is stored in a database. For example, the administrator accesses "http: / / example.com / register," the user registration screen is displayed, and the entered information is sent to the server.

[0834] Input: User information entered by the administrator on the registration screen (e.g., name, email address, role)

[0835] Output: The registered user information is saved in the database.

[0836] Step 2:

[0837] Device:

[0838] The user accesses the registration screen through the device's browser and enters the required information (name, email address, role), which is then sent to the server.

[0839] Input: Information entered by the user (e.g., name "Ichiro Tanaka", email "tanaka@example.com", role "User")

[0840] Output: The input information is sent to the server.

[0841] Step 3:

[0842] server:

[0843] The received user information is stored in a database and the user's role is determined.

[0844] Input: User information sent from the device

[0845] Output: User information is saved in the database and roles are determined.

[0846] Step 4:

[0847] User:

[0848] Specific questions about the business are entered into the terminal interface and sent to the server.

[0849] Input: Question (e.g., "How do I write a project status report?")

[0850] Output: The entered question is sent to the server.

[0851] Step 5:

[0852] Device:

[0853] The entered question is sent to the server.

[0854] Input: The question entered by the user

[0855] Output: The question is sent to the server.

[0856] Step 6:

[0857] server:

[0858] The received question is passed to the generative AI model to generate an initial answer, which is then sent back to the user from the server.

[0859] Input: User-submitted question

[0860] Output: An initial answer proposed by the generative AI model (e.g., "Project status reports should include the following elements: progress, challenges, and next steps.")

[0861] Step 7:

[0862] User:

[0863] The initial answer plan received from the server is checked on the terminal and an interface is accessed to present it to the teacher.

[0864] Input: Initial answer

[0865] Output: Display the initial answer plan in the interface to present it to the teacher.

[0866] Step 8:

[0867] Device:

[0868] Send a notification to forward the initial answer to the teacher.

[0869] Input: Initial answer

[0870] Output: Notification to forward to teacher

[0871] Step 9:

[0872] Teacher:

[0873] The user checks the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative AI model via the server.

[0874] Input: Initial answer, revisions (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action")

[0875] Output: The corrections are fed back into the generative AI model

[0876] Step 10:

[0877] server:

[0878] Based on the teacher's feedback, the generative AI model generates the final answer and provides it to the user.

[0879] Input: Teacher feedback

[0880] Output: Final proposed answer (e.g., "Project progress reports should include dates, locations, objectives, outcomes, and next actions.")

[0881] Step 11:

[0882] Device:

[0883] The final answer is received from the server and displayed to the user.

[0884] Input: Final answer

[0885] Output: Display the final proposed answer to the user

[0886] Step 12:

[0887] User:

[0888] The final answer will be confirmed and used as part of the business manual.

[0889] Input: Final answer

[0890] Output: Final answer reflected in the business manual

[0891] Step 13:

[0892] server:

[0893] Nightly batch processing is performed to update the training data and response database for the generative AI model, ensuring the accuracy and consistency of the system.

[0894] Input: Feedback information, response database

[0895] Output: Updated generative AI model and database

[0896] In this way, the system concretely performs each processing step and efficiently supports the creation and updating of business manuals.

[0897] (Application example 1)

[0898] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0899] Creating and updating operational manuals within a logistics center requires a lot of time and manpower, resulting in low work efficiency. Furthermore, as the content of operations changes daily, it is difficult to keep up with the latest information, resulting in a decline in work efficiency and an increased risk of mistakes. Therefore, a system is needed to semi-automate the creation and updating of operational manuals, thereby reducing human labor and improving work efficiency.

[0900] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0901] In this invention, the server includes means for a user to input questions related to business operations, means for receiving the questions and generating initial answer proposals using a generative model, means for presenting the initial answer proposals to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating final answers and providing them to the user, means for automatically updating the generative model and database overnight, and means for streamlining the input and generation of questions and answers specialized for logistics operations. This enables the efficient creation and updating of business manuals within a logistics center, improving work efficiency and reducing human labor.

[0902] A "user" is a person who uses the system to input questions about business and operates a terminal that receives answers.

[0903] The "teacher" is a person whose role is to check the initial answer plan and input any necessary corrections.

[0904] A "generative model" is an AI algorithm that generates initial answer proposals based on input questions and then learns and updates them based on feedback.

[0905] A "question" is an inquiry entered by a user about information or procedures related to a business.

[0906] An "initial answer proposal" is a proposal of an answer that a generative model first generates based on a question from a user.

[0907] "Correction" refers to the teacher's correction or addition to the initial answer plan.

[0908] "Feedback" is the process in which the teacher communicates the corrections to the generative model and incorporates them as training data for the model.

[0909] The "final answer" is the answer that the generative model ultimately provides after reflecting the corrections and feedback.

[0910] "Nightly update" is a process that automatically updates the generative model and database at night.

[0911] "Logistics operations" refers to all operations related to the handling, storage, shipping, etc. of goods carried out within a logistics center.

[0912] This invention is a system that streamlines the creation and updating of business manuals within a company, and is particularly applicable to operations at logistics centers. The system has three main components: a user, a teacher, and a generative model, and runs on a cloud server. Users and teachers can access the system via their smartphones.

[0913] User registration and role assignment

[0914] The server displays a user registration screen when the system is started for the first time, and provides an interface for the administrator to specify the roles of "user" and "teacher." The role information of the designated user is stored in the database.

[0915] The terminal provides a means for the user to access the registration screen and input the necessary information, which is then sent to the server.

[0916] The user enters the necessary information according to his / her role (user or teacher).

[0917] Enter a question and generate initial answer ideas

[0918] Users input questions about logistics operations into the terminal interface, and these questions are used to create an operations manual.

[0919] The terminal sends the entered question to the server, which passes it to the generative model.

[0920] The server passes the received question to the generative model to generate an initial answer, which is then sent back to the user via the device.

[0921] Teacher correction and feedback

[0922] The user checks the initial answer plan received from the server on the terminal and accesses an interface for presenting it to the teacher.

[0923] The terminal provides a means for transferring the initial answer proposal to the teacher.

[0924] The server feeds back the corrections entered by the teacher to the generative model and updates the model.

[0925] The teacher reviews the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative model via the server.

[0926] Providing a final answer

[0927] The server uses the generative model to generate a final answer based on the feedback and provides it to the user via the terminal.

[0928] The terminal displays the final answer received from the server to the user.

[0929] The user checks the final answer and uses it as content for the business manual.

[0930] Nightly updates of information

[0931] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[0932] Specific examples of operation

[0933] User Question Process:

[0934] When a user inputs a question such as "What can we do to improve the efficiency of logistics operations?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "One way to improve efficiency is to increase the speed at which products are picked."

[0935] Teacher feedback:

[0936] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, zoning the work area, using RFID tags, and providing regular training would be effective."). The server then feeds these modifications back into the generative model, updating it.

[0937] Final answer provided:

[0938] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in their work manual, saying, "To improve the efficiency of logistics operations, it is effective to zone the work area, use RFID tags, and conduct regular training."

[0939] This will enable the efficient creation and updating of business manuals within the logistics center, improving work efficiency and reducing human labor.

[0940] Example prompt sentence:

[0941] How can we improve the efficiency of picking operations at logistics centers?

[0942] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0943] Step 1:

[0944] Users input business-related questions through a smartphone application.

[0945] input:

[0946] Questions (e.g., "What can we do to improve the efficiency of our logistics operations?")

[0947] output:

[0948] Entered question data

[0949] Step 2:

[0950] The terminal sends the entered question to the server.

[0951] input:

[0952] Question data entered by the user

[0953] output:

[0954] Question data sent to the server

[0955] Step 3:

[0956] The server receives the question and passes it to the generative AI model.

[0957] input:

[0958] Question data submitted by users

[0959] output:

[0960] Question data passed to the generative AI model

[0961] Step 4:

[0962] A generative AI model generates initial answer suggestions based on the question.

[0963] input:

[0964] Question Data

[0965] Data processing:

[0966] Infer the best answer to a question using the training data inside the model

[0967] output:

[0968] Initial answer (e.g., "One way to improve efficiency is to increase the speed at which products are picked.")

[0969] Step 5:

[0970] The server receives the generated initial answer plan and transmits it to the user via the terminal.

[0971] input:

[0972] Initial draft answer

[0973] output:

[0974] The initial answer plan data is sent to the user's device.

[0975] Step 6:

[0976] The user reviews the initial answer plan and presents it to the teacher.

[0977] input:

[0978] Initial draft answer

[0979] output:

[0980] Data to present to the teacher

[0981] Step 7:

[0982] The teacher inputs corrections to the initial answer.

[0983] input:

[0984] Initial response proposal and revisions (e.g., "Specifically, zoning work areas, using RFID tags, and providing regular training are effective.")

[0985] output:

[0986] Correction data

[0987] Step 8:

[0988] The device sends the corrections made by the teacher to the server.

[0989] input:

[0990] Correction data from the teacher

[0991] output:

[0992] Modification data sent to the server

[0993] Step 9:

[0994] The server receives the corrections and feeds them back into the generative model.

[0995] input:

[0996] Correction data

[0997] output:

[0998] Data fed back into the generative model

[0999] Step 10:

[1000] The generative model learns and updates based on feedback.

[1001] input:

[1002] Feedback correction data

[1003] Data Calculation:

[1004] The process of incorporating the corrections into the training data and updating the model

[1005] output:

[1006] Updated generative model

[1007] Step 11:

[1008] The server uses the updated generative model to generate a final answer and provides it to the user.

[1009] input:

[1010] Updated generative model

[1011] Data processing:

[1012] The process of reflecting newly learned content and generating the final answer

[1013] output:

[1014] Final response data (e.g., "Zoning work areas, using RFID tags, and providing regular training are effective ways to improve the efficiency of logistics operations.")

[1015] Step 12:

[1016] The terminal displays the final answer received from the server to the user.

[1017] input:

[1018] Final response data

[1019] output:

[1020] Final response data displayed on the user's device

[1021] Step 13:

[1022] The server performs batch processing overnight to update the generative model's training data and response database to the latest information.

[1023] input:

[1024] All feedback data collected during the day

[1025] Data Calculation:

[1026] Batch processing to capture all feedback data and update the model and database

[1027] output:

[1028] Updated generative models and answer database

[1029] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1030] This invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[1031] User registration and role assignment

[1032] server:

[1033] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[1034] Device:

[1035] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[1036] User:

[1037] Enter the required information depending on your role (user or teacher).

[1038] Question input and emotion recognition

[1039] User:

[1040] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[1041] Device:

[1042] The entered question is sent to the server, and the emotion engine analyzes the user's emotional state.

[1043] server:

[1044] The system receives a question and obtains the user's emotional information analyzed by the emotion engine.Then, the question and emotional information are passed to the generative model to generate an initial answer.

[1045] Generate initial answer proposals

[1046] server:

[1047] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[1048] Device:

[1049] The initial answer plan received from the server is displayed to the user.

[1050] User:

[1051] Review the initial proposed answers.

[1052] Presentation to the teacher and provision of emotional information

[1053] User:

[1054] A request for revision is sent to present the initial answer to the teacher.

[1055] Device:

[1056] The correction request, the initial response plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[1057] server:

[1058] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[1059] Teacher feedback

[1060] Teacher:

[1061] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[1062] Device:

[1063] Enter the corrections made by the teacher and press the send button.

[1064] server:

[1065] The received corrections are fed back to the generative model to update the model.

[1066] Generate and provide the final answer

[1067] server:

[1068] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[1069] Device:

[1070] The final answer received from the server is displayed to the user.

[1071] User:

[1072] The final answer will be confirmed and used as content for the business manual.

[1073] Nightly updates of information

[1074] server:

[1075] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1076] Specific operation example

[1077] 1. User inquiry process:

[1078] When a user inputs a question such as "How should a sales report be written?", the server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to a generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[1079] 2. Teacher feedback:

[1080] The user receives the initial answer and presents it to the teacher, who then inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotion information. The server then feeds these corrections back to the generative model, updating it.

[1081] 3. Final answer provided:

[1082] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[1083] In this way, the system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1084] The processing flow will be explained below.

[1085] Program processing steps

[1086] Step 1: User registration and role assignment

[1087] server:

[1088] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[1089] Device:

[1090] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[1091] User:

[1092] Enter the required information depending on your role (user or teacher).

[1093] Step 2: Enter your question

[1094] User:

[1095] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[1096] Device:

[1097] The entered question is sent to the server.

[1098] server:

[1099] Receives a question and passes the question content to the emotion engine.

[1100] Step 3: Emotion Recognition

[1101] server:

[1102] The emotion engine analyzes the emotional state from the user's input and obtains the results. The emotion information and question are passed to the generative model.

[1103] Step 4: Generate initial answer proposals

[1104] server:

[1105] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[1106] Device:

[1107] The initial answer plan received from the server is displayed to the user.

[1108] User:

[1109] Review the initial proposed answers.

[1110] Step 5: Presenting to the teacher and providing emotional information

[1111] User:

[1112] A request for revision is sent to present the initial answer to the teacher.

[1113] Device:

[1114] The correction request, the initial answer plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[1115] server:

[1116] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[1117] Step 6: Teacher feedback

[1118] Teacher:

[1119] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[1120] Device:

[1121] Enter the corrections made by the teacher and press the send button.

[1122] server:

[1123] The received corrections are fed back to the generative model to update the model.

[1124] Step 7: Generate and deliver the final answer

[1125] server:

[1126] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[1127] Device:

[1128] The final answer received from the server is displayed to the user.

[1129] User:

[1130] The final answer will be confirmed and used as content for the business manual.

[1131] Step 8: Update information overnight

[1132] server:

[1133] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1134] Specific operation example

[1135] 1. User inquiry process:

[1136] When a user inputs a question such as "How do I write a sales report?", the device sends it to the server. The server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to the generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable for a sales report to include the following elements."

[1137] 2. Teacher feedback:

[1138] After receiving the initial answer, the user submits a correction request to present it to the teacher. The server forwards the initial answer along with the user's emotional state to the teacher. The teacher inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotional information. The server feeds these corrections back into the generative model, updating it.

[1139] 3. Final answer provided:

[1140] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[1141] As described above, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1142] Example 2

[1143] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1144] The creation and updating of business manuals has traditionally been a time-consuming and labor-intensive process, requiring a great deal of manual work. It has also been difficult to maintain consistency and accuracy in answers to questions, and there have been cases where the right answer was not given to a particular question. To solve this problem, a system is needed that can quickly and accurately update business manuals while taking into account the emotional state of the user.

[1145] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1146] In this invention, the server includes means for displaying a user registration screen when the system is first started and for users to register roles including an administrator role, means for users to input questions related to their work, means for receiving the questions and analyzing the user's emotional state with an emotion engine, and means for generating initial answer proposals with a generative model based on the questions and the analyzed emotion information. This makes it possible to provide quick and accurate answers to questions that take the user's emotional state into consideration.

[1147] A "user" is a general user of this system, who inputs questions about business.

[1148] An "administrator" is a person who has the authority to access the system and to designate users and teachers through the user registration screen when the system is started for the first time.

[1149] "Questions about business" are questions about content related to the creation and updating of business manuals within a company, and are data that are entered into the system.

[1150] An "emotion engine" is a software component that analyzes a user's emotional state in response to a question entered by the user.

[1151] A "generative model" is a machine learning algorithm that generates initial and final answers based on the input question and emotional information.

[1152] An "initial answer proposal" is an answer that is initially generated by the generative model based on the question and emotion information from the user.

[1153] The "teacher" is a person whose role is to receive the initial answer proposal from the user and make any necessary corrections.

[1154] A "request for correction" is a request that a user submits to the teacher to request correction of the initial answer plan.

[1155] The "final answer" is the answer that the generative model finally generates based on the corrections made by the teacher.

[1156] "Batch processing" is a series of processes that are executed overnight and are used to update the training data and answer database for the generative model.

[1157] The present invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[1158] Hardware and software used

[1159] This system is composed of the following hardware and software.

[1160] Server: A cloud server is used to display the user registration screen, receive questions, analyze emotions, generate suggested answers, relay correction requests, and provide final answers.

[1161] Terminal: A device used by a user, such as a computer or smartphone, that provides an interface for the user to enter questions or correction requests.

[1162] Generative model: A model that uses machine learning algorithms to generate initial and final answers based on the input question and sentiment information.

[1163] Emotion Engine: A software component that uses natural language processing techniques to parse the emotional state from a user's question.

[1164] System operation procedure

[1165] The operating procedure of this system is as follows.

[1166] 1. User Registration:

[1167] When the system is started for the first time, the server displays a user registration screen. The administrator accesses the screen via a terminal and specifies the user information and role (administrator, user, teacher). This information is saved in the database.

[1168] 2. Question input and emotion recognition:

[1169] The user inputs a business-related question into the device interface. The device then sends the question to the server, which then uses an emotion engine to analyze the user's emotional state. The analyzed emotion information is then provided to the generative model.

[1170] 3. Generate initial answer:

[1171] The server passes the question and the analyzed emotion information to the generative model to generate initial answer proposals, which are then displayed on the user's device.

[1172] 4. Request for correction:

[1173] The user checks the initial answer plan, and if they are dissatisfied, they send a correction request to the teacher. This request also includes the user's emotional information. The device sends the correction request and emotional information to the server, and the server forwards them to the teacher.

[1174] 5. Teacher feedback:

[1175] The teacher reviews the initial answer and emotion information, inputs any necessary corrections, and sends them to the server. The server then feeds the received corrections back to the generative model, which then generates the final answer.

[1176] 6. Providing a Final Response:

[1177] The generated final answer is sent from the server to the user terminal, and the user reflects the final answer in the business manual.

[1178] 7. Nightly updates:

[1179] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1180] Specific operation example

[1181] A specific example of operation is shown below.

[1182] 1. User inquiry process:

[1183] The user inputs a question such as "How do I write a sales report?" The server uses an emotion engine to analyze the user's emotional state. The question along with the analysis results is sent to the generative model, which generates an initial answer: "A sales report should include the following elements: dates, destinations, objectives, results, and next actions."

[1184] 2. Teacher feedback:

[1185] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[1186] 3. Final answer provided:

[1187] Based on the revised content, the generative model generates the final answer: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Specific examples should also be added, such as 'Date: 2023 / 10 / 01'." The server then provides the final answer to the user.

[1188] In this way, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1189] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1190] Step 1:

[1191] User registration and role assignment

[1192] Server: Display the user registration screen when the system is started for the first time.

[1193] Input: System first boot.

[1194] Specific operation: The server generates a registration screen that can be accessed on a browser and displays a "New User Registration" button.

[1195] Output: Display of registration screen.

[1196] Terminal: The administrator accesses the registration screen and enters user information and roles (user role, teacher role).

[1197] Input: Enter your user information (name, email address, password) and role.

[1198] Specific operation: The administrator enters information such as "Yamada Taro," "taro@example.com," and "Teacher role."

[1199] Output: Sending user information and roles.

[1200] Server: Receives input information and stores role information in a database.

[1201] Input: User information and role.

[1202] Specific operation: The server stores information such as "Yamada Taro," "taro@example.com," and "teacher" in a database.

[1203] Output: User information and roles are saved in the database.

[1204] Step 2:

[1205] Question input and emotion recognition

[1206] User: Enters business-related questions into the terminal interface.

[1207] Input: The question.

[1208] Specific Action: User types "How do I write a sales report?"

[1209] Output: Question data.

[1210] Terminal: Sends the entered question to the server.

[1211] Input: Question data.

[1212] Specific operation: The terminal sends the question data "How do you write a report for sales activities?" to the server.

[1213] Output: Sends the query data to the server.

[1214] Server: Analyzes the user's emotional state using the emotion engine.

[1215] Input: Question data.

[1216] Specific operation: The emotion engine determines the user's stress level as "medium" from the question.

[1217] Output: Emotional information.

[1218] Step 3:

[1219] Generate initial answer proposals

[1220] Server: Requests a generative model based on the question and emotion information to generate an initial answer.

[1221] Input: Question data and sentiment information.

[1222] Specific behavior: The generative AI model generates the following: "Sales activity reports should include the following elements: date, visit, purpose, results, and next actions."

[1223] Output: Initial draft answer.

[1224] Terminal: Show initial answer ideas to the user.

[1225] Input: Initial answer proposal.

[1226] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, destinations, objectives, results, and next actions."

[1227] Output: Display of initial answer proposal to user.

[1228] Step 4:

[1229] Submitting a revision request

[1230] User: If dissatisfied with the initial answer, send a request for revision to the teacher.

[1231] Input: Initial response proposal, requested revisions.

[1232] Specific action: The user sends a correction request with the comment "I would like more specific examples."

[1233] Output: Submit a correction request.

[1234] Terminal: Sends the initial answer proposal, including the request for revision and emotional information, to the teacher.

[1235] Input: Request for revision, initial response plan, sentiment information.

[1236] Specific action: The device sends the comment "I want more concrete examples" and the user's emotional information to the server, which then forwards it to the teacher.

[1237] Output: Transfer to teacher.

[1238] Step 5:

[1239] Teacher feedback

[1240] Teacher: Check the initial answer and emotional information provided and enter any corrections.

[1241] Input: Initial answer proposal, sentiment information, revisions.

[1242] Specific Action: The teacher types, "In your sales activity report, include the date, destination, objective, results, and next action. Also, add specific examples to each section."

[1243] Output: The corrections.

[1244] On your device: Click the Modify button to send the modifications to the server.

[1245] Input: The correction.

[1246] Specific actions: The teacher enters the corrections and clicks the submit button.

[1247] Output: Send to server.

[1248] Server: Feedback the received corrections to the generative model and update the model.

[1249] Input: The correction.

[1250] Specific operation: The server feeds the corrections back to the generative model, and the model is updated.

[1251] Output: An updated generative model.

[1252] Step 6:

[1253] Generate and provide the final answer

[1254] Server: Based on the teacher's feedback, the generative model generates the final answer.

[1255] Input: The updated generative model.

[1256] Specific behavior: The generative AI model generates, "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[1257] Output: The final answer.

[1258] Terminal: Display the final answer to the user.

[1259] Input: Final answer.

[1260] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[1261] Output: What is displayed to the user.

[1262] Step 7:

[1263] Nightly updates of information

[1264] Server: Performs batch processing overnight to update the generative model training data and answer database to the latest state.

[1265] Input: Latest fix information.

[1266] Specific operation: Batch processing is performed automatically at 11:00 PM, and the generative model adds new correction information to the training data.

[1267] Output: Updated training data and answer database.

[1268] (Application example 2)

[1269] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1270] Creating and updating manuals for factory operations requires a great deal of effort and time. Operations procedures for industrial machinery, in particular, involve many steps, and require frequent manual correction and verification, resulting in reduced efficiency and increased error rates. It is also important to ensure that employees accurately understand and follow the appropriate procedures. Failure to do so can lead to reduced productivity and potential quality issues.

[1271] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1272] In this invention, the server includes means for a user to input a question related to a business, means for receiving the question and generating an initial answer proposal using a generative model, means for presenting the initial answer proposal to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating a final answer and providing it to the user, means for providing analyzed emotion information to the generative model, and means for generating work procedures for industrial machinery and providing the work procedures to the user and the industrial machinery. This enables efficient and accurate creation and updating of work procedure manuals.

[1273] A "user" is a person who utilizes the system to enter business-related questions and review the generated answers.

[1274] The "teacher" is a person whose role is to review the initial answer proposals generated by the generative model and provide any necessary corrections or feedback.

[1275] A "generative model" is an artificial intelligence model that generates initial answers to input questions and automatically learns and updates based on feedback.

[1276] "Emotion information" is data related to emotions analyzed from questions entered by users, and is information that influences the generation of answers by the generative model.

[1277] "Industrial machinery" refers to production equipment and work robots used in factories.

[1278] "Work procedures" are guidelines that describe specific steps for carrying out a particular task or operation.

[1279] The "server" is a central computer that manages and operates the entire system, receiving questions, generating answer proposals, analyzing emotional information, and processing feedback.

[1280] The "database" is an information management system for storing and managing the generated final answers and feedback information.

[1281] "Batch processing" is a method for automatically updating and processing database information at regular intervals.

[1282] This invention is a system that enables factory robots and workers to efficiently create and update work procedures. The main components include a server, terminals, users, a teacher, a generative model, an emotion engine, and industrial machinery. This system operates as follows:

[1283] Server Features

[1284] The server is the central control and performs the following main tasks:

[1285] 1. Question reception and emotion recognition: The system receives a question entered by the user and analyzes the emotion information related to the question using an emotion engine. An emotion engine such as IBM Watson is sometimes used.

[1286] 2. Initial answer generation: The question along with the emotion information is passed to a generative model to generate an initial answer. OpenAI GPT-3 and other models are sometimes used as generative AI models.

[1287] 3. Corrective feedback: The corrective feedback from the teacher is received and fed back to the generative model. This feedback allows the generative model to learn and update automatically.

[1288] 4. Final answer generation: A final answer is generated based on the feedback and provided to the user.

[1289] 5. Information update: The final answers and related data are saved in the database and processed in daily batches.

[1290] Device Role

[1291] The terminal acts as a link between the user and the server and performs the following tasks:

[1292] 1. Providing a question input interface: Provide an interface that allows users to input questions about their work.

[1293] 2. Display answers: The initial answer proposals and final answers received from the server are displayed to the user.

[1294] 3. Sending correction request to teacher: Sends the user's correction request to the server.

[1295] User Roles

[1296] Users interact with the system as follows:

[1297] 1. Question input: A question about the business is input into the terminal and sent to the server.

[1298] 2. Confirmation of answers: Confirm the initial and final answers and request corrections if necessary.

[1299] 3. Use of the final answer: Use the final answer as an operations manual.

[1300] The role of the teacher

[1301] The teacher reviews the initial answer proposal and emotion information and provides any necessary corrections to the server, which then updates the generative model based on this information.

[1302] Usage example

[1303] For example, a user might input a question such as, "Please tell me the assembly procedure for a new product." The server uses an emotion engine to analyze the user's emotional state and passes this information along with the question to the generative model. The generative model generates an initial answer proposal, "The assembly procedure for a new product is as follows...," and provides this to the user. The user presents this initial answer proposal to the teacher, who then provides a correction such as, "Specifically, please also include the amount of force used when tightening the screws." The server feeds this correction back into the generative model, which then generates a final answer. The final answer is, "The assembly procedure for a new product is as follows...Please also include the amount of force used when tightening the screws."

[1304] Prompt Sentence Examples

[1305] Input: "What are the assembly steps for my new product?"

[1306] Initial response: "Here are the assembly steps for your new product..."

[1307] Feedback: "Please also include details on how much force to use when tightening the screws."

[1308] In this way, the system can provide efficient and accurate production and update of work procedures in industrial settings.

[1309] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1310] Step 1:

[1311] The user inputs a question about the job into the terminal. The input data is a specific question related to the job. For example, the user might input, "Please tell me the assembly procedure for a new product." The terminal then sends this input data to the server.

[1312] Step 2:

[1313] The server analyzes the question received from the user. First, it uses an emotion engine (e.g., IBM Watson) to analyze the emotional information contained in the question. Emotional information is data that indicates the user's level of urgency and importance regarding the question. This emotional information and the question content are passed to a generative AI model (e.g., OpenAI GPT-3).

[1314] Step 3:

[1315] The server generates an initial answer using a generative AI model. The generative AI model generates an appropriate answer based on the user's question and emotional information. For example, it generates an initial answer such as, "The assembly procedure for a new product is as follows..." This initial answer is returned from the server to the user via the device.

[1316] Step 4:

[1317] The user checks the initial answer provided by the server. After checking, the user inputs a correction request if necessary and sends it to the server via the terminal. The correction request includes specific corrections, such as "Specifically, please also include the amount of force used when tightening the screws."

[1318] Step 5:

[1319] The server forwards the correction request received from the user to the teacher, who then checks the correction request and the initial answer plan and enters any necessary corrections or supplementary information. The teacher's feedback is then sent back to the server.

[1320] Step 6:

[1321] The server receives corrective feedback from the teacher and sends it back to the generative AI model. This feedback updates the generative model, improving its accuracy. The generative AI model then generates a new answer based on the feedback and provides the final answer.

[1322] Step 7:

[1323] The final answer is transferred from the server to the terminal and provided to the user. For example, it might be, "The assembly procedure for a new product is as follows... Please also include the amount of force to use when tightening the screws." The user uses this final answer as their work manual.

[1324] Step 8:

[1325] At night, the server runs batch processing to update the information in the database. This keeps the generative model training data and the final answer database up to date. Updating the database is important for maintaining the accuracy and consistency of the entire system.

[1326] In this way, the server generates initial answer proposals using the user's question and emotional information, and then provides an accurate final answer after receiving feedback from a teacher.

[1327] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1328] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1329] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1330] [Third embodiment]

[1331] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1332] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1333] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1334] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1335] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1336] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1337] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1338] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1339] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1340] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1341] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1342] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1343] This invention is a system that streamlines the creation and updating of business manuals within a company, and is composed of three main components: a user, a teacher, and a generative model. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[1344] User registration and role assignment

[1345] server:

[1346] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[1347] Device:

[1348] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[1349] User:

[1350] Enter the required information depending on your role (user or teacher).

[1351] Enter a question and generate initial answer ideas

[1352] User:

[1353] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[1354] Device:

[1355] The entered question is sent to the server, which receives it.

[1356] server:

[1357] The received question is passed to a generative model to generate an initial answer, which is then sent back to the user.

[1358] Teacher correction and feedback

[1359] User:

[1360] The user checks the initial answer proposal received from the server on the terminal and accesses an interface to present it to the teacher.

[1361] Device:

[1362] Provide a method for transferring initial answer ideas to the teacher.

[1363] server:

[1364] The corrections entered by the teacher are fed back to the generative model, and the model is updated.

[1365] Teacher:

[1366] The teacher reviews the initial answer and inputs any necessary corrections, which are then fed back to the generative model via the server.

[1367] Providing a final answer

[1368] server:

[1369] Based on the feedback, the generative model generates a final answer and provides it to the user.

[1370] Device:

[1371] The final answer received from the server is displayed to the user.

[1372] User:

[1373] The final answer will be confirmed and used as part of the business manual.

[1374] Nightly updates of information

[1375] server:

[1376] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1377] Specific examples of operation

[1378] 1. User inquiry process:

[1379] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[1380] 2. Teacher feedback:

[1381] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[1382] 3. Final answer provided:

[1383] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[1384] In this way, the system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1385] The processing flow will be explained below.

[1386] Program processing steps

[1387] Step 1: User registration and role assignment

[1388] server:

[1389] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[1390] Device:

[1391] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[1392] User:

[1393] Enter the required information depending on your role (user or teacher).

[1394] Step 2: Enter your question

[1395] User:

[1396] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[1397] Device:

[1398] The entered question is sent to the server.

[1399] server:

[1400] Receives a question and passes it to a generative model.

[1401] Step 3: Generate initial answer proposals

[1402] server:

[1403] Based on the received question, the generative model generates an initial answer proposal, which is returned to the user.

[1404] Device:

[1405] The initial answer plan received from the server is displayed to the user.

[1406] User:

[1407] Review the initial proposed answers.

[1408] Step 4: Present to the teacher

[1409] User:

[1410] A request for revision is sent to present the initial answer to the teacher.

[1411] Device:

[1412] Forward the correction request and initial answer proposal to the teacher.

[1413] server:

[1414] Relays correction requests and initial response proposals to be forwarded to the teacher.

[1415] Step 5: Teacher feedback

[1416] Teacher:

[1417] Check the initial response plan, enter any necessary corrections, and send the corrections to the server.

[1418] Device:

[1419] Enter the corrections made by the teacher and press the send button.

[1420] server:

[1421] The received corrections are fed back to the generative model to update the model.

[1422] Step 6: Generate and provide the final answer

[1423] server:

[1424] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[1425] Device:

[1426] The final answer received from the server is displayed to the user.

[1427] User:

[1428] The final answer will be confirmed and used as content for the business manual.

[1429] Step 7: Nightly update of information

[1430] server:

[1431] Batch processing is performed overnight to update the generative model training data and answer database, thereby maintaining the accuracy and consistency of the entire system.

[1432] Specific operation example

[1433] 1. User inquiry process:

[1434] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[1435] 2. Teacher feedback:

[1436] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[1437] 3. Final answer provided:

[1438] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[1439] As a result, this system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1440] Example 1

[1441] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1442] Creating and updating business manuals within a company is time-consuming and labor-intensive, requiring a great deal of effort. Delays in updating information can also lead to problems with reduced business efficiency. In particular, there are limitations to the current manual process, which requires quick responses to questions while maintaining the quality of the answers. The purpose of this invention is to solve these problems and improve the efficiency of creating and updating business manuals.

[1443] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1444] In this invention, the server includes: means for a user to input a question related to a business; means for receiving the question and generating an initial answer using a generative AI model; means for the user to confirm the initial answer and present it to a teacher via a terminal; means for receiving corrections from the teacher and feeding them back to the generative AI model; means for generating a final answer and providing it to the user; and means for performing batch processing overnight to update the learning data and answer database of the generative AI model. This semi-automates the creation and updating of business manuals, significantly reducing the human labor required by companies while maintaining the accuracy and consistency of information.

[1445] A "user" is a person who operates the system and inputs questions about business.

[1446] A "terminal" is a device, such as a computer or mobile device, that a user uses to access the system.

[1447] The "server" is a central computer system that runs in a cloud environment and processes and manages questions from users and feedback from teachers.

[1448] A "generative AI model" is an artificial intelligence model that automatically generates answers to questions from users, and learns and updates based on feedback.

[1449] "Initial answer proposal" refers to the answer that the generative AI model first generates in response to a user's question.

[1450] The "teacher" is a person whose role is to review the initial proposed answers, input any necessary corrections, and provide feedback to the generative AI model.

[1451] "Final answer" refers to the answer that the generative AI model ultimately generates based on the teacher's feedback.

[1452] "Batch processing" refers to data updates and processing that are carried out in the background all at once during a certain period of time, such as at night.

[1453] A "business manual" refers to a document that describes the procedures, methods, rules, etc. for carrying out business.

[1454] A "database" is a data recording device or system for centrally managing and storing information within a system.

[1455] This invention is a system for streamlining the creation and updating of business manuals within a company. This system mainly consists of four elements: a server, a terminal, a user, and a generative AI model.

[1456] User registration and role assignment

[1457] server:

[1458] When the system is started for the first time, a user registration screen is displayed. This screen provides an interface where the administrator can designate the roles of "user" and "teacher." For example, when the administrator accesses "http: / / example.com / register," the user registration screen is displayed. The user information entered through this interface is saved in a database on the server.

[1459] Device:

[1460] The user accesses the registration screen and enters the required information (e.g., name, email address, role). For example, if the user enters the name "Ichiro Tanaka", email address "tanaka@example.com", and role "User", and clicks the send button, the information is sent to the server.

[1461] Enter a question and generate initial answer ideas

[1462] User:

[1463] Specific questions about the work are entered into the terminal interface. These questions are used to help create a work manual. For example, "How should I write a project progress report?"

[1464] Device:

[1465] The entered question is sent to the server.

[1466] server:

[1467] The received question is passed to a generative AI model, which generates an initial answer. The generated initial answer is then sent back to the user from the server. For example, if a generative AI model (e.g., GPT-4) is given the prompt, "How do I write a project progress report?", it will generate an initial answer saying, "A project progress report should include the following elements: progress, challenges, and next steps."

[1468] Teacher correction and feedback

[1469] User:

[1470] The terminal confirms the initial answer proposal received from the server and accesses an interface to present it to the teacher. For example, the user receives the answer "It is desirable to describe progress, challenges, and next steps," and transitions to a screen to send it to the teacher.

[1471] Device:

[1472] A notification will be sent to forward the initial answer to the teacher.

[1473] Teacher:

[1474] The teacher reviews the initial answer and inputs any necessary corrections. These corrections are fed back to the generative AI model via the server. For example, the teacher might correct the answer to "specifically, include the date, destination, purpose, outcome, and next action," and this is fed back to the generative AI model.

[1475] Providing a final answer

[1476] server:

[1477] Based on the teacher's feedback, the generative AI model generates a final answer and provides it to the user. For example, based on the updated feedback, the generative AI model generates a final answer such as "It is desirable for project progress reports to include dates, visits, objectives, results, and next actions."

[1478] Device:

[1479] The final answer is received from the server and displayed to the user. For example, the server may say, "It is desirable that the project progress report include the date, visit destination, objectives, results, and next actions," and the answer is displayed to the user.

[1480] Nightly updates of information

[1481] server:

[1482] Batch processing is performed at certain times during the night to update the generative AI model's learning data and answer database. This ensures the accuracy and consistency of the system. For example, the server starts batch processing every night at 2:00, updating the generative AI model's learning data with the latest feedback information.

[1483] This system streamlines the creation and updating of business manuals, reducing the amount of human effort required by companies while maintaining the accuracy and consistency of information.

[1484] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1485] Step 1:

[1486] server:

[1487] When the system is started for the first time, an interface is provided that displays a user registration screen. The administrator registers a new user and designates the role of "user" or "teacher." This registration information is stored in a database. For example, the administrator accesses "http: / / example.com / register," the user registration screen is displayed, and the entered information is sent to the server.

[1488] Input: User information entered by the administrator on the registration screen (e.g., name, email address, role)

[1489] Output: The registered user information is saved in the database.

[1490] Step 2:

[1491] Device:

[1492] The user accesses the registration screen through the device's browser and enters the required information (name, email address, role), which is then sent to the server.

[1493] Input: Information entered by the user (e.g., name "Ichiro Tanaka", email "tanaka@example.com", role "User")

[1494] Output: The input information is sent to the server.

[1495] Step 3:

[1496] server:

[1497] The received user information is stored in a database and the user's role is determined.

[1498] Input: User information sent from the device

[1499] Output: User information is saved in the database and roles are determined.

[1500] Step 4:

[1501] User:

[1502] Specific questions about the business are entered into the terminal interface and sent to the server.

[1503] Input: Question (e.g., "How do I write a project status report?")

[1504] Output: The entered question is sent to the server.

[1505] Step 5:

[1506] Device:

[1507] The entered question is sent to the server.

[1508] Input: The question entered by the user

[1509] Output: The question is sent to the server.

[1510] Step 6:

[1511] server:

[1512] The received question is passed to the generative AI model to generate an initial answer, which is then sent back to the user from the server.

[1513] Input: User-submitted question

[1514] Output: An initial answer proposed by the generative AI model (e.g., "Project status reports should include the following elements: progress, challenges, and next steps.")

[1515] Step 7:

[1516] User:

[1517] The initial answer plan received from the server is checked on the terminal and an interface is accessed to present it to the teacher.

[1518] Input: Initial answer

[1519] Output: Display the initial answer plan in the interface to present it to the teacher.

[1520] Step 8:

[1521] Device:

[1522] Send a notification to forward the initial answer to the teacher.

[1523] Input: Initial answer

[1524] Output: Notification to forward to teacher

[1525] Step 9:

[1526] Teacher:

[1527] The user checks the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative AI model via the server.

[1528] Input: Initial answer, revisions (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action")

[1529] Output: The corrections are fed back into the generative AI model

[1530] Step 10:

[1531] server:

[1532] Based on the teacher's feedback, the generative AI model generates the final answer and provides it to the user.

[1533] Input: Teacher feedback

[1534] Output: Final proposed answer (e.g., "Project progress reports should include dates, locations, objectives, outcomes, and next actions.")

[1535] Step 11:

[1536] Device:

[1537] The final answer is received from the server and displayed to the user.

[1538] Input: Final answer

[1539] Output: Display the final proposed answer to the user

[1540] Step 12:

[1541] User:

[1542] The final answer will be confirmed and used as part of the business manual.

[1543] Input: Final answer

[1544] Output: Final answer reflected in the business manual

[1545] Step 13:

[1546] server:

[1547] Nightly batch processing is performed to update the training data and response database for the generative AI model, ensuring the accuracy and consistency of the system.

[1548] Input: Feedback information, response database

[1549] Output: Updated generative AI model and database

[1550] In this way, the system concretely performs each processing step and efficiently supports the creation and updating of business manuals.

[1551] (Application example 1)

[1552] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1553] Creating and updating operational manuals within a logistics center requires a lot of time and manpower, resulting in low work efficiency. Furthermore, as the content of operations changes daily, it is difficult to keep up with the latest information, resulting in a decline in work efficiency and an increased risk of mistakes. Therefore, a system is needed to semi-automate the creation and updating of operational manuals, thereby reducing human labor and improving work efficiency.

[1554] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1555] In this invention, the server includes means for a user to input questions related to business operations, means for receiving the questions and generating initial answer proposals using a generative model, means for presenting the initial answer proposals to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating final answers and providing them to the user, means for automatically updating the generative model and database overnight, and means for streamlining the input and generation of questions and answers specialized for logistics operations. This enables the efficient creation and updating of business manuals within a logistics center, improving work efficiency and reducing human labor.

[1556] A "user" is a person who uses the system to input questions about business and operates a terminal that receives answers.

[1557] The "teacher" is a person whose role is to check the initial answer plan and input any necessary corrections.

[1558] A "generative model" is an AI algorithm that generates initial answer proposals based on input questions and then learns and updates them based on feedback.

[1559] A "question" is an inquiry entered by a user about information or procedures related to a business.

[1560] An "initial answer proposal" is a proposal of an answer that a generative model first generates based on a question from a user.

[1561] "Correction" refers to the teacher's correction or addition to the initial answer plan.

[1562] "Feedback" is the process in which the teacher communicates the corrections to the generative model and incorporates them as training data for the model.

[1563] The "final answer" is the answer that the generative model ultimately provides after reflecting the corrections and feedback.

[1564] "Nightly update" is a process that automatically updates the generative model and database at night.

[1565] "Logistics operations" refers to all operations related to the handling, storage, shipping, etc. of goods carried out within a logistics center.

[1566] This invention is a system that streamlines the creation and updating of business manuals within a company, and is particularly applicable to operations at logistics centers. The system has three main components: a user, a teacher, and a generative model, and runs on a cloud server. Users and teachers can access the system via their smartphones.

[1567] User registration and role assignment

[1568] The server displays a user registration screen when the system is started for the first time, and provides an interface for the administrator to specify the roles of "user" and "teacher." The role information of the designated user is stored in the database.

[1569] The terminal provides a means for the user to access the registration screen and input the necessary information, which is then sent to the server.

[1570] The user enters the necessary information according to his / her role (user or teacher).

[1571] Enter a question and generate initial answer ideas

[1572] Users input questions about logistics operations into the terminal interface, and these questions are used to create an operations manual.

[1573] The terminal sends the entered question to the server, which passes it to the generative model.

[1574] The server passes the received question to the generative model to generate an initial answer, which is then sent back to the user via the device.

[1575] Teacher correction and feedback

[1576] The user checks the initial answer plan received from the server on the terminal and accesses an interface for presenting it to the teacher.

[1577] The terminal provides a means for transferring the initial answer proposal to the teacher.

[1578] The server feeds back the corrections entered by the teacher to the generative model and updates the model.

[1579] The teacher reviews the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative model via the server.

[1580] Providing a final answer

[1581] The server uses the generative model to generate a final answer based on the feedback and provides it to the user via the terminal.

[1582] The terminal displays the final answer received from the server to the user.

[1583] The user checks the final answer and uses it as content for the business manual.

[1584] Nightly updates of information

[1585] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1586] Specific examples of operation

[1587] User Question Process:

[1588] When a user inputs a question such as "What can we do to improve the efficiency of logistics operations?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "One way to improve efficiency is to increase the speed at which products are picked."

[1589] Teacher feedback:

[1590] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, zoning the work area, using RFID tags, and providing regular training would be effective."). The server then feeds these modifications back into the generative model, updating it.

[1591] Final answer provided:

[1592] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in their work manual, saying, "To improve the efficiency of logistics operations, it is effective to zone the work area, use RFID tags, and conduct regular training."

[1593] This will enable the efficient creation and updating of business manuals within the logistics center, improving work efficiency and reducing human labor.

[1594] Example prompt sentence:

[1595] How can we improve the efficiency of picking operations at logistics centers?

[1596] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1597] Step 1:

[1598] Users input business-related questions through a smartphone application.

[1599] input:

[1600] Questions (e.g., "What can we do to improve the efficiency of our logistics operations?")

[1601] output:

[1602] Entered question data

[1603] Step 2:

[1604] The terminal sends the entered question to the server.

[1605] input:

[1606] Question data entered by the user

[1607] output:

[1608] Question data sent to the server

[1609] Step 3:

[1610] The server receives the question and passes it to the generative AI model.

[1611] input:

[1612] Question data submitted by users

[1613] output:

[1614] Question data passed to the generative AI model

[1615] Step 4:

[1616] A generative AI model generates initial answer suggestions based on the question.

[1617] input:

[1618] Question Data

[1619] Data processing:

[1620] Infer the best answer to a question using the training data inside the model

[1621] output:

[1622] Initial answer (e.g., "One way to improve efficiency is to increase the speed at which products are picked.")

[1623] Step 5:

[1624] The server receives the generated initial answer plan and transmits it to the user via the terminal.

[1625] input:

[1626] Initial draft answer

[1627] output:

[1628] The initial answer plan data is sent to the user's device.

[1629] Step 6:

[1630] The user reviews the initial answer plan and presents it to the teacher.

[1631] input:

[1632] Initial draft answer

[1633] output:

[1634] Data to present to the teacher

[1635] Step 7:

[1636] The teacher inputs corrections to the initial answer.

[1637] input:

[1638] Initial response proposal and revisions (e.g., "Specifically, zoning work areas, using RFID tags, and providing regular training are effective.")

[1639] output:

[1640] Correction data

[1641] Step 8:

[1642] The device sends the corrections made by the teacher to the server.

[1643] input:

[1644] Correction data from the teacher

[1645] output:

[1646] Modification data sent to the server

[1647] Step 9:

[1648] The server receives the corrections and feeds them back into the generative model.

[1649] input:

[1650] Correction data

[1651] output:

[1652] Data fed back into the generative model

[1653] Step 10:

[1654] The generative model learns and updates based on feedback.

[1655] input:

[1656] Feedback correction data

[1657] Data Calculation:

[1658] The process of incorporating the corrections into the training data and updating the model

[1659] output:

[1660] Updated generative model

[1661] Step 11:

[1662] The server uses the updated generative model to generate a final answer and provides it to the user.

[1663] input:

[1664] Updated generative model

[1665] Data processing:

[1666] The process of reflecting newly learned content and generating the final answer

[1667] output:

[1668] Final response data (e.g., "Zoning work areas, using RFID tags, and providing regular training are effective ways to improve the efficiency of logistics operations.")

[1669] Step 12:

[1670] The terminal displays the final answer received from the server to the user.

[1671] input:

[1672] Final response data

[1673] output:

[1674] Final response data displayed on the user's device

[1675] Step 13:

[1676] The server performs batch processing overnight to update the generative model's training data and response database to the latest information.

[1677] input:

[1678] All feedback data collected during the day

[1679] Data Calculation:

[1680] Batch processing to capture all feedback data and update the model and database

[1681] output:

[1682] Updated generative models and answer database

[1683] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1684] This invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[1685] User registration and role assignment

[1686] server:

[1687] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[1688] Device:

[1689] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[1690] User:

[1691] Enter the required information depending on your role (user or teacher).

[1692] Question input and emotion recognition

[1693] User:

[1694] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[1695] Device:

[1696] The entered question is sent to the server, and the emotion engine analyzes the user's emotional state.

[1697] server:

[1698] The system receives a question and obtains the user's emotional information analyzed by the emotion engine.Then, the question and emotional information are passed to the generative model to generate an initial answer.

[1699] Generate initial answer proposals

[1700] server:

[1701] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[1702] Device:

[1703] The initial answer plan received from the server is displayed to the user.

[1704] User:

[1705] Review the initial proposed answers.

[1706] Presentation to the teacher and provision of emotional information

[1707] User:

[1708] A request for revision is sent to present the initial answer to the teacher.

[1709] Device:

[1710] The correction request, the initial response plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[1711] server:

[1712] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[1713] Teacher feedback

[1714] Teacher:

[1715] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[1716] Device:

[1717] Enter the corrections made by the teacher and press the send button.

[1718] server:

[1719] The received corrections are fed back to the generative model to update the model.

[1720] Generate and provide the final answer

[1721] server:

[1722] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[1723] Device:

[1724] The final answer received from the server is displayed to the user.

[1725] User:

[1726] The final answer will be confirmed and used as content for the business manual.

[1727] Nightly updates of information

[1728] server:

[1729] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1730] Specific operation example

[1731] 1. User inquiry process:

[1732] When a user inputs a question such as "How should a sales report be written?", the server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to a generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[1733] 2. Teacher feedback:

[1734] The user receives the initial answer and presents it to the teacher, who then inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotion information. The server then feeds these corrections back to the generative model, updating it.

[1735] 3. Final answer provided:

[1736] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[1737] In this way, the system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1738] The processing flow will be explained below.

[1739] Program processing steps

[1740] Step 1: User registration and role assignment

[1741] server:

[1742] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[1743] Device:

[1744] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[1745] User:

[1746] Enter the required information depending on your role (user or teacher).

[1747] Step 2: Enter your question

[1748] User:

[1749] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[1750] Device:

[1751] The entered question is sent to the server.

[1752] server:

[1753] Receives a question and passes the question content to the emotion engine.

[1754] Step 3: Emotion Recognition

[1755] server:

[1756] The emotion engine analyzes the emotional state from the user's input and obtains the results. The emotion information and question are passed to the generative model.

[1757] Step 4: Generate initial answer proposals

[1758] server:

[1759] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[1760] Device:

[1761] The initial answer plan received from the server is displayed to the user.

[1762] User:

[1763] Review the initial proposed answers.

[1764] Step 5: Presenting to the teacher and providing emotional information

[1765] User:

[1766] A request for revision is sent to present the initial answer to the teacher.

[1767] Device:

[1768] The correction request, the initial answer plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[1769] server:

[1770] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[1771] Step 6: Teacher feedback

[1772] Teacher:

[1773] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[1774] Device:

[1775] Enter the corrections made by the teacher and press the send button.

[1776] server:

[1777] The received corrections are fed back to the generative model to update the model.

[1778] Step 7: Generate and deliver the final answer

[1779] server:

[1780] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[1781] Device:

[1782] The final answer received from the server is displayed to the user.

[1783] User:

[1784] The final answer will be confirmed and used as content for the business manual.

[1785] Step 8: Update information overnight

[1786] server:

[1787] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1788] Specific operation example

[1789] 1. User inquiry process:

[1790] When a user inputs a question such as "How do I write a sales report?", the device sends it to the server. The server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to the generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable for a sales report to include the following elements."

[1791] 2. Teacher feedback:

[1792] After receiving the initial answer, the user submits a correction request to present it to the teacher. The server forwards the initial answer along with the user's emotional state to the teacher. The teacher inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotional information. The server feeds these corrections back into the generative model, updating it.

[1793] 3. Final answer provided:

[1794] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[1795] As described above, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1796] Example 2

[1797] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1798] The creation and updating of business manuals has traditionally been a time-consuming and labor-intensive process, requiring a great deal of manual work. It has also been difficult to maintain consistency and accuracy in answers to questions, and there have been cases where the right answer was not given to a particular question. To solve this problem, a system is needed that can quickly and accurately update business manuals while taking into account the emotional state of the user.

[1799] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1800] In this invention, the server includes means for displaying a user registration screen when the system is first started and for users to register roles including an administrator role, means for users to input questions related to their work, means for receiving the questions and analyzing the user's emotional state with an emotion engine, and means for generating initial answer proposals with a generative model based on the questions and the analyzed emotion information. This makes it possible to provide quick and accurate answers to questions that take the user's emotional state into consideration.

[1801] A "user" is a general user of this system, who inputs questions about business.

[1802] An "administrator" is a person who has the authority to access the system and to designate users and teachers through the user registration screen when the system is started for the first time.

[1803] "Questions about business" are questions about content related to the creation and updating of business manuals within a company, and are data that are entered into the system.

[1804] An "emotion engine" is a software component that analyzes a user's emotional state in response to a question entered by the user.

[1805] A "generative model" is a machine learning algorithm that generates initial and final answers based on the input question and emotional information.

[1806] An "initial answer proposal" is an answer that is initially generated by the generative model based on the question and emotion information from the user.

[1807] The "teacher" is a person whose role is to receive the initial answer proposal from the user and make any necessary corrections.

[1808] A "request for correction" is a request that a user submits to the teacher to request correction of the initial answer plan.

[1809] The "final answer" is the answer that the generative model finally generates based on the corrections made by the teacher.

[1810] "Batch processing" is a series of processes that are executed overnight and are used to update the training data and answer database for the generative model.

[1811] The present invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[1812] Hardware and software used

[1813] This system is composed of the following hardware and software.

[1814] Server: A cloud server is used to display the user registration screen, receive questions, analyze emotions, generate suggested answers, relay correction requests, and provide final answers.

[1815] Terminal: A device used by a user, such as a computer or smartphone, that provides an interface for the user to enter questions or correction requests.

[1816] Generative model: A model that uses machine learning algorithms to generate initial and final answers based on the input question and sentiment information.

[1817] Emotion Engine: A software component that uses natural language processing techniques to parse the emotional state from a user's question.

[1818] System operation procedure

[1819] The operating procedure of this system is as follows.

[1820] 1. User Registration:

[1821] When the system is started for the first time, the server displays a user registration screen. The administrator accesses the screen via a terminal and specifies the user information and role (administrator, user, teacher). This information is saved in the database.

[1822] 2. Question input and emotion recognition:

[1823] The user inputs a business-related question into the device interface. The device then sends the question to the server, which then uses an emotion engine to analyze the user's emotional state. The analyzed emotion information is then provided to the generative model.

[1824] 3. Generate initial answer:

[1825] The server passes the question and the analyzed emotion information to the generative model to generate initial answer proposals, which are then displayed on the user's device.

[1826] 4. Request for correction:

[1827] The user checks the initial answer plan, and if they are dissatisfied, they send a correction request to the teacher. This request also includes the user's emotional information. The device sends the correction request and emotional information to the server, and the server forwards them to the teacher.

[1828] 5. Teacher feedback:

[1829] The teacher reviews the initial answer and emotion information, inputs any necessary corrections, and sends them to the server. The server then feeds the received corrections back to the generative model, which then generates the final answer.

[1830] 6. Providing a Final Response:

[1831] The generated final answer is sent from the server to the user terminal, and the user reflects the final answer in the business manual.

[1832] 7. Nightly updates:

[1833] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[1834] Specific operation example

[1835] A specific example of operation is shown below.

[1836] 1. User inquiry process:

[1837] The user inputs a question such as "How do I write a sales report?" The server uses an emotion engine to analyze the user's emotional state. The question along with the analysis results is sent to the generative model, which generates an initial answer: "A sales report should include the following elements: dates, destinations, objectives, results, and next actions."

[1838] 2. Teacher feedback:

[1839] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[1840] 3. Final answer provided:

[1841] Based on the revised content, the generative model generates the final answer: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Specific examples should also be added, such as 'Date: 2023 / 10 / 01'." The server then provides the final answer to the user.

[1842] In this way, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[1843] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1844] Step 1:

[1845] User registration and role assignment

[1846] Server: Display the user registration screen when the system is started for the first time.

[1847] Input: System first boot.

[1848] Specific operation: The server generates a registration screen that can be accessed on a browser and displays a "New User Registration" button.

[1849] Output: Display of registration screen.

[1850] Terminal: The administrator accesses the registration screen and enters user information and roles (user role, teacher role).

[1851] Input: Enter your user information (name, email address, password) and role.

[1852] Specific operation: The administrator enters information such as "Yamada Taro," "taro@example.com," and "Teacher role."

[1853] Output: Sending user information and roles.

[1854] Server: Receives input information and stores role information in a database.

[1855] Input: User information and role.

[1856] Specific operation: The server stores information such as "Yamada Taro," "taro@example.com," and "teacher" in a database.

[1857] Output: User information and roles are saved in the database.

[1858] Step 2:

[1859] Question input and emotion recognition

[1860] User: Enters business-related questions into the terminal interface.

[1861] Input: The question.

[1862] Specific Action: User types "How do I write a sales report?"

[1863] Output: Question data.

[1864] Terminal: Sends the entered question to the server.

[1865] Input: Question data.

[1866] Specific operation: The terminal sends the question data "How do you write a report for sales activities?" to the server.

[1867] Output: Sends the query data to the server.

[1868] Server: Analyzes the user's emotional state using the emotion engine.

[1869] Input: Question data.

[1870] Specific operation: The emotion engine determines the user's stress level as "medium" from the question.

[1871] Output: Emotional information.

[1872] Step 3:

[1873] Generate initial answer proposals

[1874] Server: Requests a generative model based on the question and emotion information to generate an initial answer.

[1875] Input: Question data and sentiment information.

[1876] Specific behavior: The generative AI model generates the following: "Sales activity reports should include the following elements: date, visit, purpose, results, and next actions."

[1877] Output: Initial draft answer.

[1878] Terminal: Show initial answer ideas to the user.

[1879] Input: Initial answer proposal.

[1880] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, destinations, objectives, results, and next actions."

[1881] Output: Display of initial answer proposal to user.

[1882] Step 4:

[1883] Submitting a revision request

[1884] User: If dissatisfied with the initial answer, send a request for revision to the teacher.

[1885] Input: Initial response proposal, requested revisions.

[1886] Specific action: The user sends a correction request with the comment "I would like more specific examples."

[1887] Output: Submit a correction request.

[1888] Terminal: Sends the initial answer proposal, including the request for revision and emotional information, to the teacher.

[1889] Input: Request for revision, initial response plan, sentiment information.

[1890] Specific action: The device sends the comment "I want more concrete examples" and the user's emotional information to the server, which then forwards it to the teacher.

[1891] Output: Transfer to teacher.

[1892] Step 5:

[1893] Teacher feedback

[1894] Teacher: Check the initial answer and emotional information provided and enter any corrections.

[1895] Input: Initial answer proposal, sentiment information, revisions.

[1896] Specific Action: The teacher types, "In your sales activity report, include the date, destination, objective, results, and next action. Also, add specific examples to each section."

[1897] Output: The corrections.

[1898] On your device: Click the Modify button to send the modifications to the server.

[1899] Input: The correction.

[1900] Specific actions: The teacher enters the corrections and clicks the submit button.

[1901] Output: Send to server.

[1902] Server: Feedback the received corrections to the generative model and update the model.

[1903] Input: The correction.

[1904] Specific operation: The server feeds the corrections back to the generative model, and the model is updated.

[1905] Output: An updated generative model.

[1906] Step 6:

[1907] Generate and provide the final answer

[1908] Server: Based on the teacher's feedback, the generative model generates the final answer.

[1909] Input: The updated generative model.

[1910] Specific behavior: The generative AI model generates, "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[1911] Output: The final answer.

[1912] Terminal: Display the final answer to the user.

[1913] Input: Final answer.

[1914] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[1915] Output: What is displayed to the user.

[1916] Step 7:

[1917] Nightly updates of information

[1918] Server: Performs batch processing overnight to update the generative model training data and answer database to the latest state.

[1919] Input: Latest fix information.

[1920] Specific operation: Batch processing is performed automatically at 11:00 PM, and the generative model adds new correction information to the training data.

[1921] Output: Updated training data and answer database.

[1922] (Application example 2)

[1923] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1924] Creating and updating manuals for factory operations requires a great deal of effort and time. Operations procedures for industrial machinery, in particular, involve many steps, and require frequent manual correction and verification, resulting in reduced efficiency and increased error rates. It is also important to ensure that employees accurately understand and follow the appropriate procedures. Failure to do so can lead to reduced productivity and potential quality issues.

[1925] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1926] In this invention, the server includes means for a user to input a question related to a business, means for receiving the question and generating an initial answer proposal using a generative model, means for presenting the initial answer proposal to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating a final answer and providing it to the user, means for providing analyzed emotion information to the generative model, and means for generating work procedures for industrial machinery and providing the work procedures to the user and the industrial machinery. This enables efficient and accurate creation and updating of work procedure manuals.

[1927] A "user" is a person who utilizes the system to enter business-related questions and review the generated answers.

[1928] The "teacher" is a person whose role is to review the initial answer proposals generated by the generative model and provide any necessary corrections or feedback.

[1929] A "generative model" is an artificial intelligence model that generates initial answers to input questions and automatically learns and updates based on feedback.

[1930] "Emotion information" is data related to emotions analyzed from questions entered by users, and is information that influences the generation of answers by the generative model.

[1931] "Industrial machinery" refers to production equipment and work robots used in factories.

[1932] "Work procedures" are guidelines that describe specific steps for carrying out a particular task or operation.

[1933] The "server" is a central computer that manages and operates the entire system, receiving questions, generating answer proposals, analyzing emotional information, and processing feedback.

[1934] The "database" is an information management system for storing and managing the generated final answers and feedback information.

[1935] "Batch processing" is a method for automatically updating and processing database information at regular intervals.

[1936] This invention is a system that enables factory robots and workers to efficiently create and update work procedures. The main components include a server, terminals, users, a teacher, a generative model, an emotion engine, and industrial machinery. This system operates as follows:

[1937] Server Features

[1938] The server is the central control and performs the following main tasks:

[1939] 1. Question reception and emotion recognition: The system receives a question entered by the user and analyzes the emotion information related to the question using an emotion engine. An emotion engine such as IBM Watson is sometimes used.

[1940] 2. Initial answer generation: The question along with the emotion information is passed to a generative model to generate an initial answer. OpenAI GPT-3 and other models are sometimes used as generative AI models.

[1941] 3. Corrective feedback: The corrective feedback from the teacher is received and fed back to the generative model. This feedback allows the generative model to learn and update automatically.

[1942] 4. Final answer generation: A final answer is generated based on the feedback and provided to the user.

[1943] 5. Information update: The final answers and related data are saved in the database and processed in daily batches.

[1944] Device Role

[1945] The terminal acts as a link between the user and the server and performs the following tasks:

[1946] 1. Providing a question input interface: Provide an interface that allows users to input questions about their work.

[1947] 2. Display answers: The initial answer proposals and final answers received from the server are displayed to the user.

[1948] 3. Sending correction request to teacher: Sends the user's correction request to the server.

[1949] User Roles

[1950] Users interact with the system as follows:

[1951] 1. Question input: A question about the business is input into the terminal and sent to the server.

[1952] 2. Confirmation of answers: Confirm the initial and final answers and request corrections if necessary.

[1953] 3. Use of the final answer: Use the final answer as an operations manual.

[1954] The role of the teacher

[1955] The teacher reviews the initial answer proposal and emotion information and provides any necessary corrections to the server, which then updates the generative model based on this information.

[1956] Usage example

[1957] For example, a user might input a question such as, "Please tell me the assembly procedure for a new product." The server uses an emotion engine to analyze the user's emotional state and passes this information along with the question to the generative model. The generative model generates an initial answer proposal, "The assembly procedure for a new product is as follows...," and provides this to the user. The user presents this initial answer proposal to the teacher, who then provides a correction such as, "Specifically, please also include the amount of force used when tightening the screws." The server feeds this correction back into the generative model, which then generates a final answer. The final answer is, "The assembly procedure for a new product is as follows...Please also include the amount of force used when tightening the screws."

[1958] Prompt Sentence Examples

[1959] Input: "What are the assembly steps for my new product?"

[1960] Initial response: "Here are the assembly steps for your new product..."

[1961] Feedback: "Please also include details on how much force to use when tightening the screws."

[1962] In this way, the system can provide efficient and accurate production and update of work procedures in industrial settings.

[1963] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1964] Step 1:

[1965] The user inputs a question about the job into the terminal. The input data is a specific question related to the job. For example, the user might input, "Please tell me the assembly procedure for a new product." The terminal then sends this input data to the server.

[1966] Step 2:

[1967] The server analyzes the question received from the user. First, it uses an emotion engine (e.g., IBM Watson) to analyze the emotional information contained in the question. Emotional information is data that indicates the user's level of urgency and importance regarding the question. This emotional information and the question content are passed to a generative AI model (e.g., OpenAI GPT-3).

[1968] Step 3:

[1969] The server generates an initial answer using a generative AI model. The generative AI model generates an appropriate answer based on the user's question and emotional information. For example, it generates an initial answer such as, "The assembly procedure for a new product is as follows..." This initial answer is returned from the server to the user via the device.

[1970] Step 4:

[1971] The user checks the initial answer provided by the server. After checking, the user inputs a correction request if necessary and sends it to the server via the terminal. The correction request includes specific corrections, such as "Specifically, please also include the amount of force used when tightening the screws."

[1972] Step 5:

[1973] The server forwards the correction request received from the user to the teacher, who then checks the correction request and the initial answer plan and enters any necessary corrections or supplementary information. The teacher's feedback is then sent back to the server.

[1974] Step 6:

[1975] The server receives corrective feedback from the teacher and sends it back to the generative AI model. This feedback updates the generative model, improving its accuracy. The generative AI model then generates a new answer based on the feedback and provides the final answer.

[1976] Step 7:

[1977] The final answer is transferred from the server to the terminal and provided to the user. For example, it might be, "The assembly procedure for a new product is as follows... Please also include the amount of force to use when tightening the screws." The user uses this final answer as their work manual.

[1978] Step 8:

[1979] At night, the server runs batch processing to update the information in the database. This keeps the generative model training data and the final answer database up to date. Updating the database is important for maintaining the accuracy and consistency of the entire system.

[1980] In this way, the server generates initial answer proposals using the user's question and emotional information, and then provides an accurate final answer after receiving feedback from a teacher.

[1981] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1982] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1983] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1984] [Fourth embodiment]

[1985] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1986] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1987] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1988] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1989] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1990] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1991] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1992] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1993] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1994] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1995] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1996] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1997] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1998] This invention is a system that streamlines the creation and updating of business manuals within a company, and is composed of three main components: a user, a teacher, and a generative model. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[1999] User registration and role assignment

[2000] server:

[2001] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[2002] Device:

[2003] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[2004] User:

[2005] Enter the required information depending on your role (user or teacher).

[2006] Enter a question and generate initial answer ideas

[2007] User:

[2008] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[2009] Device:

[2010] The entered question is sent to the server, which receives it.

[2011] server:

[2012] The received question is passed to a generative model to generate an initial answer, which is then sent back to the user.

[2013] Teacher correction and feedback

[2014] User:

[2015] The user checks the initial answer proposal received from the server on the terminal and accesses an interface to present it to the teacher.

[2016] Device:

[2017] Provide a method for transferring initial answer ideas to the teacher.

[2018] server:

[2019] The corrections entered by the teacher are fed back to the generative model, and the model is updated.

[2020] Teacher:

[2021] The teacher reviews the initial answer and inputs any necessary corrections, which are then fed back to the generative model via the server.

[2022] Providing a final answer

[2023] server:

[2024] Based on the feedback, the generative model generates a final answer and provides it to the user.

[2025] Device:

[2026] The final answer received from the server is displayed to the user.

[2027] User:

[2028] The final answer will be confirmed and used as part of the business manual.

[2029] Nightly updates of information

[2030] server:

[2031] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[2032] Specific examples of operation

[2033] 1. User inquiry process:

[2034] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[2035] 2. Teacher feedback:

[2036] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[2037] 3. Final answer provided:

[2038] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[2039] In this way, the system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[2040] The processing flow will be explained below.

[2041] Program processing steps

[2042] Step 1: User registration and role assignment

[2043] server:

[2044] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[2045] Device:

[2046] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[2047] User:

[2048] Enter the required information depending on your role (user or teacher).

[2049] Step 2: Enter your question

[2050] User:

[2051] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[2052] Device:

[2053] The entered question is sent to the server.

[2054] server:

[2055] Receives a question and passes it to a generative model.

[2056] Step 3: Generate initial answer proposals

[2057] server:

[2058] Based on the received question, the generative model generates an initial answer proposal, which is returned to the user.

[2059] Device:

[2060] The initial answer plan received from the server is displayed to the user.

[2061] User:

[2062] Review the initial proposed answers.

[2063] Step 4: Present to the teacher

[2064] User:

[2065] A request for revision is sent to present the initial answer to the teacher.

[2066] Device:

[2067] Forward the correction request and initial answer proposal to the teacher.

[2068] server:

[2069] Relays correction requests and initial response proposals to be forwarded to the teacher.

[2070] Step 5: Teacher feedback

[2071] Teacher:

[2072] Check the initial response plan, enter any necessary corrections, and send the corrections to the server.

[2073] Device:

[2074] Enter the corrections made by the teacher and press the send button.

[2075] server:

[2076] The received corrections are fed back to the generative model to update the model.

[2077] Step 6: Generate and provide the final answer

[2078] server:

[2079] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[2080] Device:

[2081] The final answer received from the server is displayed to the user.

[2082] User:

[2083] The final answer will be confirmed and used as content for the business manual.

[2084] Step 7: Nightly update of information

[2085] server:

[2086] Batch processing is performed overnight to update the generative model training data and answer database, thereby maintaining the accuracy and consistency of the entire system.

[2087] Specific operation example

[2088] 1. User inquiry process:

[2089] When a user inputs a question such as "How should a sales report be written?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[2090] 2. Teacher feedback:

[2091] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[2092] 3. Final answer provided:

[2093] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[2094] As a result, this system efficiently operates the generative model through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[2095] Example 1

[2096] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2097] Creating and updating business manuals within a company is time-consuming and labor-intensive, requiring a great deal of effort. Delays in updating information can also lead to problems with reduced business efficiency. In particular, there are limitations to the current manual process, which requires quick responses to questions while maintaining the quality of the answers. The purpose of this invention is to solve these problems and improve the efficiency of creating and updating business manuals.

[2098] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[2099] In this invention, the server includes: means for a user to input a question related to a business; means for receiving the question and generating an initial answer using a generative AI model; means for the user to confirm the initial answer and present it to a teacher via a terminal; means for receiving corrections from the teacher and feeding them back to the generative AI model; means for generating a final answer and providing it to the user; and means for performing batch processing overnight to update the learning data and answer database of the generative AI model. This semi-automates the creation and updating of business manuals, significantly reducing the human labor required by companies while maintaining the accuracy and consistency of information.

[2100] A "user" is a person who operates the system and inputs questions about business.

[2101] A "terminal" is a device, such as a computer or mobile device, that a user uses to access the system.

[2102] The "server" is a central computer system that runs in a cloud environment and processes and manages questions from users and feedback from teachers.

[2103] A "generative AI model" is an artificial intelligence model that automatically generates answers to questions from users, and learns and updates based on feedback.

[2104] "Initial answer proposal" refers to the answer that the generative AI model first generates in response to a user's question.

[2105] The "teacher" is a person whose role is to review the initial proposed answers, input any necessary corrections, and provide feedback to the generative AI model.

[2106] "Final answer" refers to the answer that the generative AI model ultimately generates based on the teacher's feedback.

[2107] "Batch processing" refers to data updates and processing that are carried out in the background all at once during a certain period of time, such as at night.

[2108] A "business manual" refers to a document that describes the procedures, methods, rules, etc. for carrying out business.

[2109] A "database" is a data recording device or system for centrally managing and storing information within a system.

[2110] This invention is a system for streamlining the creation and updating of business manuals within a company. This system mainly consists of four elements: a server, a terminal, a user, and a generative AI model.

[2111] User registration and role assignment

[2112] server:

[2113] When the system is started for the first time, a user registration screen is displayed. This screen provides an interface where the administrator can designate the roles of "user" and "teacher." For example, when the administrator accesses "http: / / example.com / register," the user registration screen is displayed. The user information entered through this interface is saved in a database on the server.

[2114] Device:

[2115] The user accesses the registration screen and enters the required information (e.g., name, email address, role). For example, if the user enters the name "Ichiro Tanaka", email address "tanaka@example.com", and role "User", and clicks the send button, the information is sent to the server.

[2116] Enter a question and generate initial answer ideas

[2117] User:

[2118] Specific questions about the work are entered into the terminal interface. These questions are used to help create a work manual. For example, "How should I write a project progress report?"

[2119] Device:

[2120] The entered question is sent to the server.

[2121] server:

[2122] The received question is passed to a generative AI model, which generates an initial answer. The generated initial answer is then sent back to the user from the server. For example, if a generative AI model (e.g., GPT-4) is given the prompt, "How do I write a project progress report?", it will generate an initial answer saying, "A project progress report should include the following elements: progress, challenges, and next steps."

[2123] Teacher correction and feedback

[2124] User:

[2125] The terminal confirms the initial answer proposal received from the server and accesses an interface to present it to the teacher. For example, the user receives the answer "It is desirable to describe progress, challenges, and next steps," and transitions to a screen to send it to the teacher.

[2126] Device:

[2127] A notification will be sent to forward the initial answer to the teacher.

[2128] Teacher:

[2129] The teacher reviews the initial answer and inputs any necessary corrections. These corrections are fed back to the generative AI model via the server. For example, the teacher might correct the answer to "specifically, include the date, destination, purpose, outcome, and next action," and this is fed back to the generative AI model.

[2130] Providing a final answer

[2131] server:

[2132] Based on the teacher's feedback, the generative AI model generates a final answer and provides it to the user. For example, based on the updated feedback, the generative AI model generates a final answer such as "It is desirable for project progress reports to include dates, visits, objectives, results, and next actions."

[2133] Device:

[2134] The final answer is received from the server and displayed to the user. For example, the server may say, "It is desirable that the project progress report include the date, visit destination, objectives, results, and next actions," and the answer is displayed to the user.

[2135] Nightly updates of information

[2136] server:

[2137] Batch processing is performed at certain times during the night to update the generative AI model's learning data and answer database. This ensures the accuracy and consistency of the system. For example, the server starts batch processing every night at 2:00, updating the generative AI model's learning data with the latest feedback information.

[2138] This system streamlines the creation and updating of business manuals, reducing the amount of human effort required by companies while maintaining the accuracy and consistency of information.

[2139] The flow of the identification process in the first embodiment will be described with reference to FIG.

[2140] Step 1:

[2141] server:

[2142] When the system is started for the first time, an interface is provided that displays a user registration screen. The administrator registers a new user and designates the role of "user" or "teacher." This registration information is stored in a database. For example, the administrator accesses "http: / / example.com / register," the user registration screen is displayed, and the entered information is sent to the server.

[2143] Input: User information entered by the administrator on the registration screen (e.g., name, email address, role)

[2144] Output: The registered user information is saved in the database.

[2145] Step 2:

[2146] Device:

[2147] The user accesses the registration screen through the device's browser and enters the required information (name, email address, role), which is then sent to the server.

[2148] Input: Information entered by the user (e.g., name "Ichiro Tanaka", email "tanaka@example.com", role "User")

[2149] Output: The input information is sent to the server.

[2150] Step 3:

[2151] server:

[2152] The received user information is stored in a database and the user's role is determined.

[2153] Input: User information sent from the device

[2154] Output: User information is saved in the database and roles are determined.

[2155] Step 4:

[2156] User:

[2157] Specific questions about the business are entered into the terminal interface and sent to the server.

[2158] Input: Question (e.g., "How do I write a project status report?")

[2159] Output: The entered question is sent to the server.

[2160] Step 5:

[2161] Device:

[2162] The entered question is sent to the server.

[2163] Input: The question entered by the user

[2164] Output: The question is sent to the server.

[2165] Step 6:

[2166] server:

[2167] The received question is passed to the generative AI model to generate an initial answer, which is then sent back to the user from the server.

[2168] Input: User-submitted question

[2169] Output: An initial answer proposed by the generative AI model (e.g., "Project status reports should include the following elements: progress, challenges, and next steps.")

[2170] Step 7:

[2171] User:

[2172] The initial answer plan received from the server is checked on the terminal and an interface is accessed to present it to the teacher.

[2173] Input: Initial answer

[2174] Output: Display the initial answer plan in the interface to present it to the teacher.

[2175] Step 8:

[2176] Device:

[2177] Send a notification to forward the initial answer to the teacher.

[2178] Input: Initial answer

[2179] Output: Notification to forward to teacher

[2180] Step 9:

[2181] Teacher:

[2182] The user checks the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative AI model via the server.

[2183] Input: Initial answer, revisions (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action")

[2184] Output: The corrections are fed back into the generative AI model

[2185] Step 10:

[2186] server:

[2187] Based on the teacher's feedback, the generative AI model generates the final answer and provides it to the user.

[2188] Input: Teacher feedback

[2189] Output: Final proposed answer (e.g., "Project progress reports should include dates, locations, objectives, outcomes, and next actions.")

[2190] Step 11:

[2191] Device:

[2192] The final answer is received from the server and displayed to the user.

[2193] Input: Final answer

[2194] Output: Display the final proposed answer to the user

[2195] Step 12:

[2196] User:

[2197] The final answer will be confirmed and used as part of the business manual.

[2198] Input: Final answer

[2199] Output: Final answer reflected in the business manual

[2200] Step 13:

[2201] server:

[2202] Nightly batch processing is performed to update the training data and response database for the generative AI model, ensuring the accuracy and consistency of the system.

[2203] Input: Feedback information, response database

[2204] Output: Updated generative AI model and database

[2205] In this way, the system concretely performs each processing step and efficiently supports the creation and updating of business manuals.

[2206] (Application example 1)

[2207] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2208] Creating and updating operational manuals within a logistics center requires a lot of time and manpower, resulting in low work efficiency. Furthermore, as the content of operations changes daily, it is difficult to keep up with the latest information, resulting in a decline in work efficiency and an increased risk of mistakes. Therefore, a system is needed to semi-automate the creation and updating of operational manuals, thereby reducing human labor and improving work efficiency.

[2209] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[2210] In this invention, the server includes means for a user to input questions related to business operations, means for receiving the questions and generating initial answer proposals using a generative model, means for presenting the initial answer proposals to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating final answers and providing them to the user, means for automatically updating the generative model and database overnight, and means for streamlining the input and generation of questions and answers specialized for logistics operations. This enables the efficient creation and updating of business manuals within a logistics center, improving work efficiency and reducing human labor.

[2211] A "user" is a person who uses the system to input questions about business and operates a terminal that receives answers.

[2212] The "teacher" is a person whose role is to check the initial answer plan and input any necessary corrections.

[2213] A "generative model" is an AI algorithm that generates initial answer proposals based on input questions and then learns and updates them based on feedback.

[2214] A "question" is an inquiry entered by a user about information or procedures related to a business.

[2215] An "initial answer proposal" is a proposal of an answer that a generative model first generates based on a question from a user.

[2216] "Correction" refers to the teacher's correction or addition to the initial answer plan.

[2217] "Feedback" is the process in which the teacher communicates the corrections to the generative model and incorporates them as training data for the model.

[2218] The "final answer" is the answer that the generative model ultimately provides after reflecting the corrections and feedback.

[2219] "Nightly update" is a process that automatically updates the generative model and database at night.

[2220] "Logistics operations" refers to all operations related to the handling, storage, shipping, etc. of goods carried out within a logistics center.

[2221] This invention is a system that streamlines the creation and updating of business manuals within a company, and is particularly applicable to operations at logistics centers. The system has three main components: a user, a teacher, and a generative model, and runs on a cloud server. Users and teachers can access the system via their smartphones.

[2222] User registration and role assignment

[2223] The server displays a user registration screen when the system is started for the first time, and provides an interface for the administrator to specify the roles of "user" and "teacher." The role information of the designated user is stored in the database.

[2224] The terminal provides a means for the user to access the registration screen and input the necessary information, which is then sent to the server.

[2225] The user enters the necessary information according to his / her role (user or teacher).

[2226] Enter a question and generate initial answer ideas

[2227] Users input questions about logistics operations into the terminal interface, and these questions are used to create an operations manual.

[2228] The terminal sends the entered question to the server, which passes it to the generative model.

[2229] The server passes the received question to the generative model to generate an initial answer, which is then sent back to the user via the device.

[2230] Teacher correction and feedback

[2231] The user checks the initial answer plan received from the server on the terminal and accesses an interface for presenting it to the teacher.

[2232] The terminal provides a means for transferring the initial answer proposal to the teacher.

[2233] The server feeds back the corrections entered by the teacher to the generative model and updates the model.

[2234] The teacher reviews the initial answer proposal and inputs any necessary corrections, which are then fed back to the generative model via the server.

[2235] Providing a final answer

[2236] The server uses the generative model to generate a final answer based on the feedback and provides it to the user via the terminal.

[2237] The terminal displays the final answer received from the server to the user.

[2238] The user checks the final answer and uses it as content for the business manual.

[2239] Nightly updates of information

[2240] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[2241] Specific examples of operation

[2242] User Question Process:

[2243] When a user inputs a question such as "What can we do to improve the efficiency of logistics operations?", the server passes the question to the generative model and generates an initial answer. The generated initial answer is returned to the user as "One way to improve efficiency is to increase the speed at which products are picked."

[2244] Teacher feedback:

[2245] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, zoning the work area, using RFID tags, and providing regular training would be effective."). The server then feeds these modifications back into the generative model, updating it.

[2246] Final answer provided:

[2247] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in their work manual, saying, "To improve the efficiency of logistics operations, it is effective to zone the work area, use RFID tags, and conduct regular training."

[2248] This will enable the efficient creation and updating of business manuals within the logistics center, improving work efficiency and reducing human labor.

[2249] Example prompt sentence:

[2250] How can we improve the efficiency of picking operations at logistics centers?

[2251] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[2252] Step 1:

[2253] Users input business-related questions through a smartphone application.

[2254] input:

[2255] Questions (e.g., "What can we do to improve the efficiency of our logistics operations?")

[2256] output:

[2257] Entered question data

[2258] Step 2:

[2259] The terminal sends the entered question to the server.

[2260] input:

[2261] Question data entered by the user

[2262] output:

[2263] Question data sent to the server

[2264] Step 3:

[2265] The server receives the question and passes it to the generative AI model.

[2266] input:

[2267] Question data submitted by users

[2268] output:

[2269] Question data passed to the generative AI model

[2270] Step 4:

[2271] A generative AI model generates initial answer suggestions based on the question.

[2272] input:

[2273] Question Data

[2274] Data processing:

[2275] Infer the best answer to a question using the training data inside the model

[2276] output:

[2277] Initial answer (e.g., "One way to improve efficiency is to increase the speed at which products are picked.")

[2278] Step 5:

[2279] The server receives the generated initial answer plan and transmits it to the user via the terminal.

[2280] input:

[2281] Initial draft answer

[2282] output:

[2283] The initial answer plan data is sent to the user's device.

[2284] Step 6:

[2285] The user reviews the initial answer plan and presents it to the teacher.

[2286] input:

[2287] Initial draft answer

[2288] output:

[2289] Data to present to the teacher

[2290] Step 7:

[2291] The teacher inputs corrections to the initial answer.

[2292] input:

[2293] Initial response proposal and revisions (e.g., "Specifically, zoning work areas, using RFID tags, and providing regular training are effective.")

[2294] output:

[2295] Correction data

[2296] Step 8:

[2297] The device sends the corrections made by the teacher to the server.

[2298] input:

[2299] Correction data from the teacher

[2300] output:

[2301] Modification data sent to the server

[2302] Step 9:

[2303] The server receives the corrections and feeds them back into the generative model.

[2304] input:

[2305] Correction data

[2306] output:

[2307] Data fed back into the generative model

[2308] Step 10:

[2309] The generative model learns and updates based on feedback.

[2310] input:

[2311] Feedback correction data

[2312] Data Calculation:

[2313] The process of incorporating the corrections into the training data and updating the model

[2314] output:

[2315] Updated generative model

[2316] Step 11:

[2317] The server uses the updated generative model to generate a final answer and provides it to the user.

[2318] input:

[2319] Updated generative model

[2320] Data processing:

[2321] The process of reflecting newly learned content and generating the final answer

[2322] output:

[2323] Final response data (e.g., "Zoning work areas, using RFID tags, and providing regular training are effective ways to improve the efficiency of logistics operations.")

[2324] Step 12:

[2325] The terminal displays the final answer received from the server to the user.

[2326] input:

[2327] Final response data

[2328] output:

[2329] Final response data displayed on the user's device

[2330] Step 13:

[2331] The server performs batch processing overnight to update the generative model's training data and response database to the latest information.

[2332] input:

[2333] All feedback data collected during the day

[2334] Data Calculation:

[2335] Batch processing to capture all feedback data and update the model and database

[2336] output:

[2337] Updated generative models and answer database

[2338] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[2339] This invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[2340] User registration and role assignment

[2341] server:

[2342] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[2343] Device:

[2344] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[2345] User:

[2346] Enter the required information depending on your role (user or teacher).

[2347] Question input and emotion recognition

[2348] User:

[2349] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[2350] Device:

[2351] The entered question is sent to the server, and the emotion engine analyzes the user's emotional state.

[2352] server:

[2353] The system receives a question and obtains the user's emotional information analyzed by the emotion engine.Then, the question and emotional information are passed to the generative model to generate an initial answer.

[2354] Generate initial answer proposals

[2355] server:

[2356] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[2357] Device:

[2358] The initial answer plan received from the server is displayed to the user.

[2359] User:

[2360] Review the initial proposed answers.

[2361] Presentation to the teacher and provision of emotional information

[2362] User:

[2363] A request for revision is sent to present the initial answer to the teacher.

[2364] Device:

[2365] The correction request, the initial response plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[2366] server:

[2367] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[2368] Teacher feedback

[2369] Teacher:

[2370] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[2371] Device:

[2372] Enter the corrections made by the teacher and press the send button.

[2373] server:

[2374] The received corrections are fed back to the generative model to update the model.

[2375] Generate and provide the final answer

[2376] server:

[2377] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[2378] Device:

[2379] The final answer received from the server is displayed to the user.

[2380] User:

[2381] The final answer will be confirmed and used as content for the business manual.

[2382] Nightly updates of information

[2383] server:

[2384] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[2385] Specific operation example

[2386] 1. User inquiry process:

[2387] When a user inputs a question such as "How should a sales report be written?", the server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to a generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable that a sales report include the following elements."

[2388] 2. Teacher feedback:

[2389] The user receives the initial answer and presents it to the teacher, who then inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotion information. The server then feeds these corrections back to the generative model, updating it.

[2390] 3. Final answer provided:

[2391] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[2392] In this way, the system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[2393] The processing flow will be explained below.

[2394] Program processing steps

[2395] Step 1: User registration and role assignment

[2396] server:

[2397] When the system is started for the first time, a user registration screen is displayed, providing an interface for the administrator to designate the roles of "user" and "teacher." The role information of the designated users is stored in the database.

[2398] Device:

[2399] The user accesses the registration screen and enters the necessary information, which is then sent to the server.

[2400] User:

[2401] Enter the required information depending on your role (user or teacher).

[2402] Step 2: Enter your question

[2403] User:

[2404] Questions about the job are entered into the terminal interface, and these questions are used to help create the job manual.

[2405] Device:

[2406] The entered question is sent to the server.

[2407] server:

[2408] Receives a question and passes the question content to the emotion engine.

[2409] Step 3: Emotion Recognition

[2410] server:

[2411] The emotion engine analyzes the emotional state from the user's input and obtains the results. The emotion information and question are passed to the generative model.

[2412] Step 4: Generate initial answer proposals

[2413] server:

[2414] Based on the question and emotion information, the generative model generates an initial answer, which is returned to the user.

[2415] Device:

[2416] The initial answer plan received from the server is displayed to the user.

[2417] User:

[2418] Review the initial proposed answers.

[2419] Step 5: Presenting to the teacher and providing emotional information

[2420] User:

[2421] A request for revision is sent to present the initial answer to the teacher.

[2422] Device:

[2423] The correction request, the initial answer plan, and the user's emotional state information analyzed by the emotion engine are forwarded to the teacher.

[2424] server:

[2425] Relays an initial response plan including a request for revision and emotional information to be forwarded to the teacher.

[2426] Step 6: Teacher feedback

[2427] Teacher:

[2428] The initial answer plan and the user's emotional state information are checked, and any necessary corrections are input. The corrections are then sent to the server.

[2429] Device:

[2430] Enter the corrections made by the teacher and press the send button.

[2431] server:

[2432] The received corrections are fed back to the generative model to update the model.

[2433] Step 7: Generate and deliver the final answer

[2434] server:

[2435] Based on the teacher's feedback, the generative model generates a final answer, which is then sent to the user.

[2436] Device:

[2437] The final answer received from the server is displayed to the user.

[2438] User:

[2439] The final answer will be confirmed and used as content for the business manual.

[2440] Step 8: Update information overnight

[2441] server:

[2442] Batch processing is performed overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[2443] Specific operation example

[2444] 1. User inquiry process:

[2445] When a user inputs a question such as "How do I write a sales report?", the device sends it to the server. The server uses an emotion engine to analyze the user's emotional state. The question and the analysis results are sent to the generative model, which generates an initial answer. The generated initial answer is returned to the user as "It is desirable for a sales report to include the following elements."

[2446] 2. Teacher feedback:

[2447] After receiving the initial answer, the user submits a correction request to present it to the teacher. The server forwards the initial answer along with the user's emotional state to the teacher. The teacher inputs the necessary corrections (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action") and the corrections based on the emotional information. The server feeds these corrections back into the generative model, updating it.

[2448] 3. Final answer provided:

[2449] The generative model generates a final answer based on the revised content, and the server provides the final answer to the user. The user reflects this in the business manual, noting that "it is desirable for sales activity reports to include the date, visit destination, purpose, results, and next actions."

[2450] As described above, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[2451] Example 2

[2452] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2453] The creation and updating of business manuals has traditionally been a time-consuming and labor-intensive process, requiring a great deal of manual work. It has also been difficult to maintain consistency and accuracy in answers to questions, and there have been cases where the right answer was not given to a particular question. To solve this problem, a system is needed that can quickly and accurately update business manuals while taking into account the emotional state of the user.

[2454] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[2455] In this invention, the server includes means for displaying a user registration screen when the system is first started and for users to register roles including an administrator role, means for users to input questions related to their work, means for receiving the questions and analyzing the user's emotional state with an emotion engine, and means for generating initial answer proposals with a generative model based on the questions and the analyzed emotion information. This makes it possible to provide quick and accurate answers to questions that take the user's emotional state into consideration.

[2456] A "user" is a general user of this system, who inputs questions about business.

[2457] An "administrator" is a person who has the authority to access the system and to designate users and teachers through the user registration screen when the system is started for the first time.

[2458] "Questions about business" are questions about content related to the creation and updating of business manuals within a company, and are data that are entered into the system.

[2459] An "emotion engine" is a software component that analyzes a user's emotional state in response to a question entered by the user.

[2460] A "generative model" is a machine learning algorithm that generates initial and final answers based on the input question and emotional information.

[2461] An "initial answer proposal" is an answer that is initially generated by the generative model based on the question and emotion information from the user.

[2462] The "teacher" is a person whose role is to receive the initial answer proposal from the user and make any necessary corrections.

[2463] A "request for correction" is a request that a user submits to the teacher to request correction of the initial answer plan.

[2464] The "final answer" is the answer that the generative model finally generates based on the corrections made by the teacher.

[2465] "Batch processing" is a series of processes that are executed overnight and are used to update the training data and answer database for the generative model.

[2466] The present invention is a system for streamlining the creation and updating of business manuals within a company, and includes a user, a teacher, a generative model, and an emotion engine. This system runs on a cloud server and can be accessed by users and teachers via their terminals.

[2467] Hardware and software used

[2468] This system is composed of the following hardware and software.

[2469] Server: A cloud server is used to display the user registration screen, receive questions, analyze emotions, generate suggested answers, relay correction requests, and provide final answers.

[2470] Terminal: A device used by a user, such as a computer or smartphone, that provides an interface for the user to enter questions or correction requests.

[2471] Generative model: A model that uses machine learning algorithms to generate initial and final answers based on the input question and sentiment information.

[2472] Emotion Engine: A software component that uses natural language processing techniques to parse the emotional state from a user's question.

[2473] System operation procedure

[2474] The operating procedure of this system is as follows.

[2475] 1. User Registration:

[2476] When the system is started for the first time, the server displays a user registration screen. The administrator accesses the screen via a terminal and specifies the user information and role (administrator, user, teacher). This information is saved in the database.

[2477] 2. Question input and emotion recognition:

[2478] The user inputs a business-related question into the device interface. The device then sends the question to the server, which then uses an emotion engine to analyze the user's emotional state. The analyzed emotion information is then provided to the generative model.

[2479] 3. Generate initial answer:

[2480] The server passes the question and the analyzed emotion information to the generative model to generate initial answer proposals, which are then displayed on the user's device.

[2481] 4. Request for correction:

[2482] The user checks the initial answer plan, and if they are dissatisfied, they send a correction request to the teacher. This request also includes the user's emotional information. The device sends the correction request and emotional information to the server, and the server forwards them to the teacher.

[2483] 5. Teacher feedback:

[2484] The teacher reviews the initial answer and emotion information, inputs any necessary corrections, and sends them to the server. The server then feeds the received corrections back to the generative model, which then generates the final answer.

[2485] 6. Providing a Final Response:

[2486] The generated final answer is sent from the server to the user terminal, and the user reflects the final answer in the business manual.

[2487] 7. Nightly updates:

[2488] The server performs batch processing overnight to update the generative model training data and answer database to the latest state, thereby maintaining the accuracy and consistency of the entire system.

[2489] Specific operation example

[2490] A specific example of operation is shown below.

[2491] 1. User inquiry process:

[2492] The user inputs a question such as "How do I write a sales report?" The server uses an emotion engine to analyze the user's emotional state. The question along with the analysis results is sent to the generative model, which generates an initial answer: "A sales report should include the following elements: dates, destinations, objectives, results, and next actions."

[2493] 2. Teacher feedback:

[2494] The user receives the initial answer and presents it to the teacher, who then inputs any necessary modifications (e.g., "Specifically, please include the date, destination, purpose, outcome, and next action"). The server then feeds these modifications back into the generative model, updating it.

[2495] 3. Final answer provided:

[2496] Based on the revised content, the generative model generates the final answer: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Specific examples should also be added, such as 'Date: 2023 / 10 / 01'." The server then provides the final answer to the user.

[2497] In this way, this system efficiently operates the generative model and emotion engine through the cooperation of users and teachers, semi-automating the creation and updating of business manuals and significantly reducing the human effort required by companies.

[2498] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2499] Step 1:

[2500] User registration and role assignment

[2501] Server: Display the user registration screen when the system is started for the first time.

[2502] Input: System first boot.

[2503] Specific operation: The server generates a registration screen that can be accessed on a browser and displays a "New User Registration" button.

[2504] Output: Display of registration screen.

[2505] Terminal: The administrator accesses the registration screen and enters user information and roles (user role, teacher role).

[2506] Input: Enter your user information (name, email address, password) and role.

[2507] Specific operation: The administrator enters information such as "Yamada Taro," "taro@example.com," and "Teacher role."

[2508] Output: Sending user information and roles.

[2509] Server: Receives input information and stores role information in a database.

[2510] Input: User information and role.

[2511] Specific operation: The server stores information such as "Yamada Taro," "taro@example.com," and "teacher" in a database.

[2512] Output: User information and roles are saved in the database.

[2513] Step 2:

[2514] Question input and emotion recognition

[2515] User: Enters business-related questions into the terminal interface.

[2516] Input: The question.

[2517] Specific Action: User types "How do I write a sales report?"

[2518] Output: Question data.

[2519] Terminal: Sends the entered question to the server.

[2520] Input: Question data.

[2521] Specific operation: The terminal sends the question data "How do you write a report for sales activities?" to the server.

[2522] Output: Sends the query data to the server.

[2523] Server: Analyzes the user's emotional state using the emotion engine.

[2524] Input: Question data.

[2525] Specific operation: The emotion engine determines the user's stress level as "medium" from the question.

[2526] Output: Emotional information.

[2527] Step 3:

[2528] Generate initial answer proposals

[2529] Server: Requests a generative model based on the question and emotion information to generate an initial answer.

[2530] Input: Question data and sentiment information.

[2531] Specific behavior: The generative AI model generates the following: "Sales activity reports should include the following elements: date, visit, purpose, results, and next actions."

[2532] Output: Initial draft answer.

[2533] Terminal: Show initial answer ideas to the user.

[2534] Input: Initial answer proposal.

[2535] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, destinations, objectives, results, and next actions."

[2536] Output: Display of initial answer proposal to user.

[2537] Step 4:

[2538] Submitting a revision request

[2539] User: If dissatisfied with the initial answer, send a request for revision to the teacher.

[2540] Input: Initial response proposal, requested revisions.

[2541] Specific action: The user sends a correction request with the comment "I would like more specific examples."

[2542] Output: Submit a correction request.

[2543] Terminal: Sends the initial answer proposal, including the request for revision and emotional information, to the teacher.

[2544] Input: Request for revision, initial response plan, sentiment information.

[2545] Specific action: The device sends the comment "I want more concrete examples" and the user's emotional information to the server, which then forwards it to the teacher.

[2546] Output: Transfer to teacher.

[2547] Step 5:

[2548] Teacher feedback

[2549] Teacher: Check the initial answer and emotional information provided and enter any corrections.

[2550] Input: Initial answer proposal, sentiment information, revisions.

[2551] Specific Action: The teacher types, "In your sales activity report, include the date, destination, objective, results, and next action. Also, add specific examples to each section."

[2552] Output: The corrections.

[2553] On your device: Click the Modify button to send the modifications to the server.

[2554] Input: The correction.

[2555] Specific actions: The teacher enters the corrections and clicks the submit button.

[2556] Output: Send to server.

[2557] Server: Feedback the received corrections to the generative model and update the model.

[2558] Input: The correction.

[2559] Specific operation: The server feeds the corrections back to the generative model, and the model is updated.

[2560] Output: An updated generative model.

[2561] Step 6:

[2562] Generate and provide the final answer

[2563] Server: Based on the teacher's feedback, the generative model generates the final answer.

[2564] Input: The updated generative model.

[2565] Specific behavior: The generative AI model generates, "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[2566] Output: The final answer.

[2567] Terminal: Display the final answer to the user.

[2568] Input: Final answer.

[2569] Specific Action: The user sees the following on their screen: "Sales activity reports should include the following elements: dates, visits, objectives, results, and next actions. Also, add specific examples, such as 'Date: 2023 / 10 / 01'."

[2570] Output: What is displayed to the user.

[2571] Step 7:

[2572] Nightly updates of information

[2573] Server: Performs batch processing overnight to update the generative model training data and answer database to the latest state.

[2574] Input: Latest fix information.

[2575] Specific operation: Batch processing is performed automatically at 11:00 PM, and the generative model adds new correction information to the training data.

[2576] Output: Updated training data and answer database.

[2577] (Application example 2)

[2578] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2579] Creating and updating manuals for factory operations requires a great deal of effort and time. Operations procedures for industrial machinery, in particular, involve many steps, and require frequent manual correction and verification, resulting in reduced efficiency and increased error rates. It is also important to ensure that employees accurately understand and follow the appropriate procedures. Failure to do so can lead to reduced productivity and potential quality issues.

[2580] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[2581] In this invention, the server includes means for a user to input a question related to a business, means for receiving the question and generating an initial answer proposal using a generative model, means for presenting the initial answer proposal to a teacher, means for receiving corrections from the teacher and feeding them back to the generative model, means for generating a final answer and providing it to the user, means for providing analyzed emotion information to the generative model, and means for generating work procedures for industrial machinery and providing the work procedures to the user and the industrial machinery. This enables efficient and accurate creation and updating of work procedure manuals.

[2582] A "user" is a person who utilizes the system to enter business-related questions and review the generated answers.

[2583] The "teacher" is a person whose role is to review the initial answer proposals generated by the generative model and provide any necessary corrections or feedback.

[2584] A "generative model" is an artificial intelligence model that generates initial answers to input questions and automatically learns and updates based on feedback.

[2585] "Emotion information" is data related to emotions analyzed from questions entered by users, and is information that influences the generation of answers by the generative model.

[2586] "Industrial machinery" refers to production equipment and work robots used in factories.

[2587] "Work procedures" are guidelines that describe specific steps for carrying out a particular task or operation.

[2588] The "server" is a central computer that manages and operates the entire system, receiving questions, generating answer proposals, analyzing emotional information, and processing feedback.

[2589] The "database" is an information management system for storing and managing the generated final answers and feedback information.

[2590] "Batch processing" is a method for automatically updating and processing database information at regular intervals.

[2591] This invention is a system that enables factory robots and workers to efficiently create and update work procedures. The main components include a server, terminals, users, a teacher, a generative model, an emotion engine, and industrial machinery. This system operates as follows:

[2592] Server Features

[2593] The server is the central control and performs the following main tasks:

[2594] 1. Question reception and emotion recognition: The system receives a question entered by the user and analyzes the emotion information related to the question using an emotion engine. An emotion engine such as IBM Watson is sometimes used.

[2595] 2. Initial answer generation: The question along with the emotion information is passed to a generative model to generate an initial answer. OpenAI GPT-3 and other models are sometimes used as generative AI models.

[2596] 3. Corrective feedback: The corrective feedback from the teacher is received and fed back to the generative model. This feedback allows the generative model to learn and update automatically.

[2597] 4. Final answer generation: A final answer is generated based on the feedback and provided to the user.

[2598] 5. Information update: The final answers and related data are saved in the database and processed in daily batches.

[2599] Device Role

[2600] The terminal acts as a link between the user and the server and performs the following tasks:

[2601] 1. Providing a question input interface: Provide an interface that allows users to input questions about their work.

[2602] 2. Display answers: The initial answer proposals and final answers received from the server are displayed to the user.

[2603] 3. Sending correction request to teacher: Sends the user's correction request to the server.

[2604] User Roles

[2605] Users interact with the system as follows:

[2606] 1. Question input: A question about the business is input into the terminal and sent to the server.

[2607] 2. Confirmation of answers: Confirm the initial and final answers and request corrections if necessary.

[2608] 3. Use of the final answer: Use the final answer as an operations manual.

[2609] The role of the teacher

[2610] The teacher reviews the initial answer proposal and emotion information and provides any necessary corrections to the server, which then updates the generative model based on this information.

[2611] Usage example

[2612] For example, a user might input a question such as, "Please tell me the assembly procedure for a new product." The server uses an emotion engine to analyze the user's emotional state and passes this information along with the question to the generative model. The generative model generates an initial answer proposal, "The assembly procedure for a new product is as follows...," and provides this to the user. The user presents this initial answer proposal to the teacher, who then provides a correction such as, "Specifically, please also include the amount of force used when tightening the screws." The server feeds this correction back into the generative model, which then generates a final answer. The final answer is, "The assembly procedure for a new product is as follows...Please also include the amount of force used when tightening the screws."

[2613] Prompt Sentence Examples

[2614] Input: "What are the assembly steps for my new product?"

[2615] Initial response: "Here are the assembly steps for your new product..."

[2616] Feedback: "Please also include details on how much force to use when tightening the screws."

[2617] In this way, the system can provide efficient and accurate production and update of work procedures in industrial settings.

[2618] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2619] Step 1:

[2620] The user inputs a question about the job into the terminal. The input data is a specific question related to the job. For example, the user might input, "Please tell me the assembly procedure for a new product." The terminal then sends this input data to the server.

[2621] Step 2:

[2622] The server analyzes the question received from the user. First, it uses an emotion engine (e.g., IBM Watson) to analyze the emotional information contained in the question. Emotional information is data that indicates the user's level of urgency and importance regarding the question. This emotional information and the question content are passed to a generative AI model (e.g., OpenAI GPT-3).

[2623] Step 3:

[2624] The server generates an initial answer using a generative AI model. The generative AI model generates an appropriate answer based on the user's question and emotional information. For example, it generates an initial answer such as, "The assembly procedure for a new product is as follows..." This initial answer is returned from the server to the user via the device.

[2625] Step 4:

[2626] The user checks the initial answer provided by the server. After checking, the user inputs a correction request if necessary and sends it to the server via the terminal. The correction request includes specific corrections, such as "Specifically, please also include the amount of force used when tightening the screws."

[2627] Step 5:

[2628] The server forwards the correction request received from the user to the teacher, who then checks the correction request and the initial answer plan and enters any necessary corrections or supplementary information. The teacher's feedback is then sent back to the server.

[2629] Step 6:

[2630] The server receives corrective feedback from the teacher and sends it back to the generative AI model. This feedback updates the generative model, improving its accuracy. The generative AI model then generates a new answer based on the feedback and provides the final answer.

[2631] Step 7:

[2632] The final answer is transferred from the server to the terminal and provided to the user. For example, it might be, "The assembly procedure for a new product is as follows... Please also include the amount of force to use when tightening the screws." The user uses this final answer as their work manual.

[2633] Step 8:

[2634] At night, the server runs batch processing to update the information in the database. This keeps the generative model training data and the final answer database up to date. Updating the database is important for maintaining the accuracy and consistency of the entire system.

[2635] In this way, the server generates initial answer proposals using the user's question and emotional information, and then provides an accurate final answer after receiving feedback from a teacher.

[2636] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2637] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[2638] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2639] The emotion identification model 59 as an emotion engine may determine the user's emo...

Claims

1. a means for a user to input a business question; means for receiving the question and generating an initial answer proposal using a generative model; means for presenting the initial answer plan to a teacher; A means of receiving corrections from the teacher and feeding them back into the generative model; means for generating and providing a final answer to the user; A system including:

2. 10. The system of claim 1, A system that includes the ability for the generative model to automatically update and learn based on feedback.

3. 10. The system of claim 1, The system includes a means for storing the final answers in a database and updating them through daily batch processing.

Images