system

Abstract

This system allows even inexperienced crew members to quickly and appropriately obtain necessary manual information, thereby improving the quality of customer service. [Solution] A system including means for reading connection settings and performing initialization; means for receiving authentication information from a user and performing authentication; means for receiving keywords from a user and processing search requests; means for searching for relevant documents from a database; means for generating new information using a generative AI if no relevant documents are found in the database; means for formatting the generated information and sending it to a user terminal; means for receiving feedback from a user and storing it in a database; and means for retraining the generative AI using the received feedback.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of the chatbot's 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

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the in-store sales business, it is particularly difficult for inexperienced staff to grasp where the manuals are and what they contain, and there is a lack of means to quickly check the appropriate manuals. As a result, the response of the staff may be delayed, and customer satisfaction may decrease. Therefore, there is a need for a system that allows staff to easily and quickly obtain the necessary manual information.

Means for Solving the Problems

[0005] The present invention provides a system that includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and sending it to a user terminal, means for receiving feedback from a user and storing it in a database, and means for retraining the generative AI using the received feedback. This makes it possible for even inexperienced crew members to quickly and appropriately obtain the necessary manual information and improve the quality of customer service.

[0006] "Connection settings" refers to a configuration file containing initialization information necessary for the system to function correctly.

[0007] "Initialization" refers to the process of loading the resources necessary when the system starts up, and includes preparatory work such as configuring AI models and database connections.

[0008] "Authentication information" refers to the information necessary to verify a user's identity, and typically includes a username and password.

[0009] A "keyword" is a word or phrase that a user enters into a system to search for specific information.

[0010] A "search request" is a request sent to a system to retrieve specific information based on keywords entered by the user.

[0011] A "database" is a system that organizes, stores, and keeps related information and documents in a searchable format.

[0012] "Documents" refer to documents and manuals stored within the system, which are sources of information that crew members refer to when performing their duties.

[0013] "Generative AI" refers to artificial intelligence models that generate new text or content based on input information.

[0014] "Formatting" is the process of converting generated information into a format that is easy to see and read.

[0015] A "user terminal" is a device used by the crew, where search requests are entered and results are displayed.

[0016] "Feedback" is the act of a user sending an evaluation or opinion about the information provided to the system.

[0017] "Retraining" refers to the learning process used to improve the performance of a generative AI based on the feedback it receives. [Brief explanation of the drawing]

[0018] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9]Shows an emotion map to which a plurality of emotions are mapped. [Figure 10] Shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Modes for Carrying Out the Invention

[0019] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0020] First, the terms used in the following description will be described.

[0021] In the following embodiments, a processor with a reference numeral (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0022] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0023] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0024] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0026] [First Embodiment]

[0027] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0028] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0031] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0034] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0038] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0039] This invention relates to a system that uses generative AI to provide manuals for in-store sales operations to crew members. This system consists of multiple elements that work in conjunction with each other. The main components of the system and their specific operations are described below.

[0040] Initialization and User Authentication

[0041] The server reads the necessary connection settings when the system starts up and initializes the AI ​​model and database connection. When a user accesses the system, the server receives the user's authentication information (username, password) from the terminal and verifies it against the database. For users who are successfully authenticated, a unique session ID is generated and returned to the terminal.

[0042] Manual search request

[0043] The user enters a question or keyword into an input field on the terminal. For example, they might enter the keyword "cash register operation new registration". The terminal then sends this keyword to the server in JSON format.

[0044] Searching and generating manuals

[0045] The server analyzes the received keywords and searches the database for relevant documents. For example, if there is a manual corresponding to the keywords "cash register operation new registration," it retrieves its contents. If no relevant documents are found, the server uses generative AI to generate new manual content. The generated content is then formatted for easier reading.

[0046] Information return and display

[0047] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button," might be displayed.

[0048] Gathering feedback

[0049] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server. The server stores the feedback in a database and uses it to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[0050] Specific example

[0051] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[0052] 2. Manual Search: The user enters "cash register operation new registration," and the terminal sends the keywords to the server. The server searches the database and either retrieves the document or generates new manual content using generative AI. The generated information is formatted and sent to the terminal for display to the user.

[0053] 3. Feedback Collection: Users input feedback such as "Helpful" regarding the displayed information, and the device sends this to the server. The server stores the feedback in a database and uses it to retrain the generative AI.

[0054] This system allows even inexperienced crew members to quickly and appropriately obtain necessary manual information, thereby improving the quality of customer service.

[0055] The following describes the processing flow.

[0056] Step 1:

[0057] When the system starts up, the server reads the configuration file and initializes the AI ​​model and database connection. This makes the system's internal resources available.

[0058] Step 2:

[0059] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[0060] Step 3:

[0061] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[0062] Step 4:

[0063] The user enters a question or keyword into the search field on the terminal. For example, they might enter "cash register operation new registration".

[0064] Step 5:

[0065] The terminal sends the keywords entered by the user to the server in JSON format. An example request is {"query": "Register Operation New Registration", "session_id": "xyz123"}.

[0066] Step 6:

[0067] The server analyzes the received search request and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[0068] Step 7:

[0069] If the server cannot find relevant documents, it invokes a generative AI to generate new information. The generated information is specific and in a format that answers the user's questions.

[0070] Step 8:

[0071] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response.

[0072] Step 9:

[0073] The terminal displays the received manual information on its screen. For example, it might show detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[0074] Step 10:

[0075] Users provide feedback on the displayed information. For example, they might rate it as "This information was helpful."

[0076] Step 11:

[0077] The device sends user feedback information to the server in JSON format. An example request is {"manual_id": "abc123", "feedback": "Helpful", "session_id": "xyz123"}.

[0078] Step 12:

[0079] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI.

[0080] Step 13:

[0081] The server uses accumulated feedback to retrain the AI ​​model to improve the performance of the generative AI. This process improves the accuracy of responses to subsequent search requests.

[0082] (Example 1)

[0083] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0084] In retail sales operations, there is a problem where crew members cannot quickly and appropriately obtain information due to outdated or incomplete manuals. Furthermore, there are issues where manuals cannot be found, or newly generated information is insufficient, leading to a decline in the quality of customer service. Additionally, there is a problem where the quality of generated information does not improve because feedback on the usefulness of the information provided is not effectively utilized.

[0085] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0086] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI model if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in the database, means for retraining the generative AI model using the received feedback, means for initializing the AI ​​model and database connection when the system starts up, means for the user to enter keywords in a search field and the terminal to send those keywords to the server, and means for formatting the generated manual information and displaying it to the user. As a result, crew members can quickly obtain the latest and most appropriate manual information, improving the quality of customer service. Furthermore, it becomes possible to continuously improve the quality of the information generated by utilizing feedback.

[0087] "Connection settings" refers to a configuration file that initializes the system and contains connection information for necessary external resources and services.

[0088] "Initialization" refers to the process of preparing the necessary settings and resources when a system starts operating.

[0089] "Authentication information" refers to information such as usernames and passwords used to identify and authenticate users.

[0090] A "keyword" refers to a phrase or term that a user enters into a search field to search for specific information or manuals.

[0091] A "search request" refers to a request that a user sends to obtain specific information.

[0092] A "database" refers to a system that systematically stores information and data, and allows for searching and retrieval as needed.

[0093] "Document" refers to a piece of paper containing specific information, such as a manual or guidelines.

[0094] A "generative AI model" refers to an artificial intelligence model that generates new text or documents based on input information.

[0095] "Formatting" refers to adjusting the format of generated information to make it easier to see and read.

[0096] A "user terminal" refers to a device used by a user to input information or to check results.

[0097] "Feedback" refers to users' evaluations and opinions on the information provided.

[0098] "Retraining" refers to the process of improving the performance of an existing generative AI model by having it retrained based on user feedback.

[0099] "Submitting keywords" refers to the act of sending search terms entered by the user to the server.

[0100] "Formatting and displaying manual information to the user" refers to the process of converting generated or acquired manual information into a visually easy-to-read format and displaying it on the user's device.

[0101] This invention relates to a system that uses a generative AI model to provide manuals for in-store sales operations to crew members. The specific processing details and embodiments of the system are described below.

[0102] Initialization and User Authentication

[0103] The server reads a connection configuration file (e.g., config.yaml) when the system starts up and initializes connections to the AI ​​model (e.g., GPT-4®) and the database (e.g., MySQL®). This process prepares the AI ​​model and database for the system.

[0104] When a user accesses the system, the terminal receives authentication information (username and password) and sends it to the server in JSON format. The server compares the authentication information with the user table in the database, and if authentication is successful, generates a unique session ID and sends it back to the terminal.

[0105] Manual search request

[0106] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. The terminal sends this keyword to the server in JSON format.

[0107] Searching and generating manuals

[0108] The server analyzes the received keywords and uses a language analysis library (e.g., NLTK) to evaluate their relevance. Next, it searches the database for relevant manual documents. If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. The generated information is then formatted visually using a Python formatting library.

[0109] Information return and display

[0110] The server sends formatted manual information to the terminal in JSON format. The terminal converts the received data into a viewable format and displays it to the user. For example, it might display specific instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[0111] Gathering feedback

[0112] Users evaluate whether the information provided was helpful and enter feedback. For example, a rating such as "This information was helpful." The device sends this feedback to the server in JSON format. The server stores the received feedback in a database. This feedback is used to retrain generative AI models. Retraining is done to improve the model's response accuracy so that responses to subsequent search requests are more accurate.

[0113] Specific example

[0114] When a user enters their username and password on the login screen, the terminal sends this information to the server. If authentication is successful, the server generates a session ID and sends it back to the terminal. Next, the user enters keywords such as "cash register operation new registration," and the terminal sends these to the server. The server searches the database and either retrieves relevant documents or generates a new manual using a generative AI model. Finally, the terminal displays the generated information to the user, who enters feedback and sends it back to the system. Based on this feedback, the generative AI model is retrained.

[0115] Example of a prompt

[0116] "Register a new user for register operations"

[0117] "Procedure for registering new products at the register"

[0118] In this way, users can obtain the necessary information quickly and accurately. This can improve the efficiency of in-store sales operations and the quality of customer service.

[0119] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0120] Step 1:

[0121] Initialization

[0122] The server reads a connection configuration file (e.g., config.yaml) when the system starts up. It receives the path to the configuration file as input and sets the connection string and model parameters. Based on this, it initializes the connection to the AI ​​model (e.g., GPT-4) and the database (e.g., MySQL). It generates a log confirming the successful connection as output.

[0123] Step 2:

[0124] Receiving user authentication information

[0125] The user accesses the login screen and enters their username and password. The terminal receives this input information in JSON format and sends it to the server. This passes the authentication information to the server. The input consists of the username and password, and the output is JSON data containing the authentication information sent to the server.

[0126] Step 3:

[0127] Verification of authentication information

[0128] The server compares the received authentication information with the user table in the database. It uses the username and password as input to query the database and retrieve matching records. As output, if authentication is successful, it generates a unique session ID and sends it back to the terminal. If authentication fails, it returns an error message.

[0129] Step 4:

[0130] Keyword Input

[0131] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. This issues a search request to the system. The keyword entered by the user is passed to the terminal as input. As output, the search query data in JSON format is sent from the terminal to the server.

[0132] Step 5:

[0133] Keyword analysis and database search

[0134] The server extracts keywords from the received JSON data and parses them using a language analysis library (e.g., NLTK). The input is keywords, which are parsed to generate conditions for searching related documents. The output is a generated database query, which is then sent to the database.

[0135] Step 6:

[0136] Generating new manuals using generative AI models

[0137] If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. Keyword-based prompts (e.g., "new product register registration procedure") are sent to the AI ​​model as input. The generated text information is returned as output.

[0138] Step 7:

[0139] Information formatting

[0140] The server formats the generated text information to make it visually readable. The input is the result generated by an AI model, and formatting adjustments are made using a Python formatting library. The output is the formatted manual information.

[0141] Step 8:

[0142] Sending manual information

[0143] The server sends the formatted manual information to the terminal in JSON format. The input is the formatted manual information, and the output is a data packet generated to send this information to the terminal.

[0144] Step 9:

[0145] Display of manual information

[0146] The terminal converts the received manual information into a displayable format and displays it to the user. The input is manual information in JSON format, which is then rendered as HTML and UI components. The output is a visually displayed manual provided to the user.

[0147] Step 10:

[0148] Feedback Input

[0149] Users provide feedback on the information provided. For example, they rate the information they found helpful. The input is the feedback information, and the output is this information sent from the terminal to the server.

[0150] Step 11:

[0151] Save feedback

[0152] The server saves the received feedback to the database. The input is the feedback information, and the output is a log that confirms the save was successful.

[0153] Step 12:

[0154] Retraining generative AI models

[0155] The server retrains the generative AI model based on the feedback. The feedback data is used as input, serving as the dataset for training the AI ​​model. The output is a new model with improved performance.

[0156] (Application Example 1)

[0157] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0158] In autonomous vehicles, a challenge exists in that drivers and passengers often struggle to obtain quick and accurate information regarding driving operations and system settings when manuals are required. This problem is particularly pronounced in emergencies or when using an autonomous vehicle for the first time. Current information provision methods make it difficult to provide appropriate manuals in real time, which could compromise safety and convenience.

[0159] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0160] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and transmitting it to a user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, and means for providing manuals regarding driving operations and system settings on information terminals of autonomous vehicles and personal mobile terminals. This enables drivers and passengers to quickly and accurately obtain information regarding the operation of autonomous vehicles.

[0161] "Connection settings" refer to the specific settings that enable each component of a system to communicate with one another.

[0162] "Initialization" refers to the process of setting up the state of each component when a system starts up, preparing it for normal operation.

[0163] "Authentication information" refers to data used to verify a user's identity, and typically includes a username and password.

[0164] "Authentication" refers to the process of verifying whether a user has legitimate access rights based on the authentication information provided.

[0165] A "keyword" refers to a word or phrase that a user enters to search for specific information.

[0166] A "search request" refers to a request sent by a user to search for specific information.

[0167] A "database" refers to a system that systematically stores information so that it can be searched and retrieved later.

[0168] A "document" refers to a document that describes specific information or procedures.

[0169] "Generative AI" refers to a type of artificial intelligence that has the ability to generate new information and documents.

[0170] A "manual" refers to a document that explains specific operating procedures or system settings.

[0171] "Formatting" refers to the process of converting generated information into a format that is easy to see and understand.

[0172] A "user terminal" refers to a device used by a user to input information or view results. Examples include smartphones and vehicle information terminals.

[0173] "Feedback" refers to opinions, evaluations, and input regarding the effectiveness of information provided by users.

[0174] "Retraining" refers to the process of improving the performance of an existing AI model by having it learn again using new data.

[0175] An "autonomous vehicle" refers to a vehicle that has the ability to drive automatically without requiring direct control by a driver.

[0176] An "information terminal" refers to a device installed inside an autonomous vehicle that allows the user to input and confirm information.

[0177] The system of this invention is designed to quickly provide manuals regarding driving operations and system settings to information terminals in autonomous vehicles and to the driver's personal mobile device. The detailed processing steps and hardware and software used to implement this system are described below.

[0178] 1. Initialization and User Authentication

[0179] The server reads the necessary connection settings when the system starts up and initializes the database connection and generative AI model. When a user accesses the system, the server receives user authentication information (username, password) from the terminal and authenticates the user by comparing it with the database. For users who successfully authenticate, a unique session ID is generated and returned to the terminal. The hardware required here is a server and a personal mobile device, and the software utilizes a database system (e.g., MongoDB or SQL) and a generative AI model.

[0180] 2. Manual search request

[0181] The user enters a question or keyword into an input field on the device. For example, they might enter the keyword "How to turn on autonomous driving mode." The device sends this keyword to the server in JSON format. The server analyzes the received keyword and searches its database for relevant documents. The software used here utilizes a web framework (e.g., Flask).

[0182] 3. Searching for and generating manuals

[0183] The server searches the database for relevant documents based on the analyzed keywords. If no relevant documents are found, the server uses a generative AI model to generate new manual content. The generated content is then formatted for readability. The specific AI library used in this process is Some AI Library.

[0184] 4. Information return and display

[0185] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel," are displayed. Information terminals and smartphones are used as hardware for this purpose.

[0186] 5. Gathering feedback and retraining

[0187] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server, which stores it in a database. Furthermore, this feedback is used to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[0188] Examples of specific cases and prompt statements

[0189] As a concrete example, imagine a driver who wants to know "how to turn on autonomous driving mode" and inputs "Please tell me how to turn on autonomous driving mode." The system responds to this input as follows:

[0190] To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel.

[0191] In this way, users can quickly and accurately obtain information regarding the operation of autonomous vehicles.

[0192] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0193] Step 1:

[0194] The server reads the connection settings when the system starts up and performs database connection and generative AI model initialization. Specifically, the server reads DB connection information from the configuration file and establishes a connection to the database. It also reads the generative AI model from the specified path and completes its initialization. The input is the configuration file, and the output is the status of successful database connection and AI model initialization.

[0195] Step 2:

[0196] The user enters authentication information (username, password) from the terminal, and the terminal sends this information to the server. The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal. The input is the user's authentication information, and the output is the session ID or the authentication error status. Specifically, the server compares the authentication information with the user information in the database and generates a session ID if the match is successful.

[0197] Step 3:

[0198] The user enters a question or keyword into an input field on the device, and the device sends this keyword to the server in JSON format. The input is the keyword entered by the user, and the output is a search request in JSON format. Specifically, the device acquires the user's input in real time, formats it into JSON format, and sends it to the server.

[0199] Step 4:

[0200] The server parses the received keywords and searches the database for relevant documents. The input is a search request in JSON format, and the output is a flag indicating whether relevant documents or search results are not found. Specifically, the server creates a database query based on the keywords and performs a database search.

[0201] Step 5:

[0202] If no relevant documents are found in the database, the server uses a generative AI model to generate new manual content. The generated content is formatted and converted into a readable format. The input consists of flags and keywords indicating that no search results were found, and the output is the generated manual content. Specifically, the server inputs keywords as prompts to the generative AI model and receives the text generated by the AI ​​model. It then formats it into the appropriate format.

[0203] Step 6:

[0204] The server sends formatted manual information to the terminal, which then displays it to the user. The input is the formatted manual content, and the output is the specific operating procedure displayed to the user. Specifically, the server sends the manual content to the terminal, which receives it and displays it visually to the user.

[0205] Step 7:

[0206] The user evaluates whether the provided information was helpful and enters feedback. The terminal sends this feedback to the server. The input is the user's feedback, and the output is the transmission of the feedback data to the server. Specifically, the terminal retrieves the user's rating, formats it in JSON format, and sends it to the server.

[0207] Step 8:

[0208] The server stores the received feedback in a database and uses it to retrain the generative AI model. The input is the feedback data, and the output is the feedback stored in the database and a new dataset used for retraining. Specifically, the server stores the feedback in the database and periodically retrieves and uses the feedback data to retrain the generative AI model.

[0209] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0210] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system operates through a process from login to user emotion recognition, manual generation, feedback collection, and AI retraining. The main components and specific operation of the system are described below.

[0211] Initialization and User Authentication

[0212] When the system starts, the server loads the connection settings and initializes the AI ​​model, emotion engine, and database connection. When a user accesses the system, the terminal enters a username and password. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[0213] Manual search request

[0214] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format.

[0215] Emotion recognition by an emotion engine

[0216] At the same time, the emotion engine analyzes the user's facial expressions and voice during input to recognize their emotional state. For example, if a user makes an unhappy facial expression or makes an unhappy voice during input, the emotion engine analyzes this and determines that they are "unhappy."

[0217] Searching and generating manuals

[0218] The server analyzes the received search request and searches the database for relevant documents. If relevant documents are found, their contents are retrieved. If no documents are found, a generative AI is used to generate new manual content. Based on the analysis results of the emotion engine, the tone and content of the generated manual are adjusted.

[0219] Information return and display

[0220] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific instructions such as "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button" are displayed. If the information is deemed inappropriate, additional support information or actions (e.g., "View more detailed guide" or "Contact support center") are suggested.

[0221] Gathering feedback

[0222] Users provide feedback on the information provided, for example, by rating it as "This information was very helpful." The device sends the feedback information to the server, which may include the results of the sentiment engine's analysis.

[0223] Saving feedback and retraining the AI

[0224] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI model and the emotion engine, improving the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[0225] Specific example

[0226] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[0227] 2. Manual Search and Emotion Recognition: The user enters "cash register operation new registration," and the terminal sends the keyword along with facial expressions and voice data to the server. The server searches the database and, if necessary, generates new manual content using generative AI. The generated information reflects the analysis results of the emotion engine and is formatted in an appropriate tone and style. The formatted information is then displayed to the user through the terminal.

[0228] 3. Feedback Collection and Retraining: Users provide feedback such as "This was very helpful" regarding the information, and the device sends this to the server along with the analysis results of the emotion engine. The server stores the feedback and retrains the generative AI and emotion engine.

[0229] This system allows crew members to quickly and appropriately obtain necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state.

[0230] The following describes the processing flow.

[0231] Step 1:

[0232] When the system starts up, the server loads the connection settings and initializes the connections for the AI ​​model, emotion engine, and database. This makes the system's internal resources available.

[0233] Step 2:

[0234] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[0235] Step 3:

[0236] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[0237] Step 4:

[0238] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format. An example request is {"query": "cash register operation new registration", "session_id": "xyz123"}.

[0239] Step 5:

[0240] Simultaneously, the emotion engine collects the user's facial expressions and voice during input and analyzes the user's emotional state. For example, if the user's facial expression indicates displeasure, the emotion engine recognizes this as "unpleasant" and sends the analysis result to the server.

[0241] Step 6:

[0242] The server analyzes the received search request and sentiment state, and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[0243] Step 7:

[0244] If the server cannot find the relevant document in the database, it invokes a generative AI to generate new manual content. The generated content is then adjusted to an appropriate tone and content based on the analysis results of the emotion engine. For example, if the user is in an unpleasant state, the information will be generated using gentler and more polite language.

[0245] Step 8:

[0246] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response. For example, the response may also include additional support information tailored to the user's emotional state.

[0247] Step 9:

[0248] The terminal displays the received manual information on the screen. For example, it displays detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button." If the information is deemed inappropriate, additional support information (e.g., a link to a detailed guide or a direct contact number for support) is suggested.

[0249] Step 10:

[0250] Users provide feedback on the displayed information. For example, they might rate it as "This information was very helpful."

[0251] Step 11:

[0252] The device sends user feedback information to the server in JSON format. This information may include the results of the emotion engine's analysis. An example request is {"manual_id": "abc123", "feedback": "Helpful", "emotion": "happy", "session_id": "xyz123"}.

[0253] Step 12:

[0254] The server stores the received feedback information in a database and uses it to retrain the emotion engine and generative AI models. This data is used to improve the analysis accuracy and response quality of the AI ​​and emotion engine.

[0255] Step 13:

[0256] The server retrains its generative AI model and emotion engine based on accumulated feedback and emotion data. This improves the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[0257] This process allows crew members to quickly and appropriately obtain necessary manual information, as well as respond flexibly to the user's emotional state, thereby improving operational efficiency and customer satisfaction.

[0258] (Example 2)

[0259] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0260] In modern retail operations, manuals are needed to enable crew members to quickly and accurately obtain information. However, existing systems struggle to respond flexibly, taking into account the emotional state of users, and search results can sometimes be inaccurate. Furthermore, AI is not being retrained based on feedback, resulting in insufficient system performance improvements. A system is needed to address these challenges.

[0261] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0262] In this invention, the server includes means for reading and initializing connection settings; means for receiving and authenticating authentication information from the user; means for receiving keywords from the user and processing search requests; means for analyzing the user's facial expressions and voice during input and recognizing their emotional state; means for searching for relevant documents from a database; means for generating new information using generative AI if no relevant documents are found in the database; means for formatting the generated information based on the user's emotional state and sending it to the user terminal; means for receiving feedback from the user and storing it in a database; and means for retraining the generative AI using the received feedback. As a result, crew members can quickly and appropriately obtain the necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state. Furthermore, the accuracy of the AI ​​model can be continuously improved by utilizing the feedback.

[0263] "Connection settings" refer to the settings that include network and database connection information necessary for servers and terminals to communicate.

[0264] "Initialization" refers to the process of loading the necessary settings and data when a system starts up, and preparing the system to be operational.

[0265] "Authentication information" refers to usernames, passwords, and other authentication methods used to identify a user and grant them access.

[0266] A "search request" is a set of keywords or queries that a user sends to a server to retrieve specific information.

[0267] "Emotional state" refers to the state of emotions analyzed from the user's facial expressions and voice, and means feelings such as discomfort or joy.

[0268] "Generative AI" refers to artificial intelligence technology that has the ability to generate new information based on large amounts of data. Examples include natural language generation models.

[0269] A "user terminal" refers to a device, such as a computer or smartphone, that a user uses to access and operate a system.

[0270] "Feedback" refers to evaluations and opinions provided by users of a system, and is used to improve the system and enhance its performance.

[0271] "Retraining" is the process of using newly acquired data to train an artificial intelligence model again and improve its performance.

[0272] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system includes a series of processes from user login to emotion recognition, manual generation, feedback collection, and AI retraining.

[0273] Initialization and User Authentication

[0274] The server reads the connection settings when the system starts up and initializes the AI ​​model, emotion engine, and database connection. The specific software used is a Python®-based Flask server, and the generated AI model uses either TENSORFLOW® or PyTorch. The emotion engine uses OpenFace or EmotionAPI. The user accesses the system and enters a username and password on the login screen. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[0275] Manual search request

[0276] The user enters a keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal converts this keyword into JSON format and sends it to the server.

[0277] Emotion recognition by an emotion engine

[0278] During this process, the emotion engine analyzes the user's facial expressions and voice. For example, if the user makes an unhappy facial expression or makes an unspoken voice, the emotion engine will determine this to be "unpleasant." OpenCV and TensorFlow models are used to collect and analyze the device's camera and microphone data in real time for emotion analysis.

[0279] Searching and generating manuals

[0280] The server analyzes the received search request and searches the database for relevant documents. If no matching documents are found in the database, it uses a generative AI (e.g., GPT-3®) to generate new manual content. Based on the results of the sentiment engine's analysis, the tone and content of the generated manual are adjusted. For example, if dissatisfaction is identified, detailed explanations and additional support information will be included.

[0281] Information return and display

[0282] The formatted manual information is sent from the server to the terminal, and the terminal displays it to the user. For example, specific operation procedures such as "To register a new product at the cash register, scan the product barcode, enter the price and quantity, and then click the registration button" are displayed. If it is determined to be unpleasant, additional support information and actions (e.g., "Display more detailed guide" or "Contact the support center") are proposed.

[0283] Collection of Feedback

[0284] The user inputs feedback on the provided information. For example, evaluate it as "This information was very helpful". Since the feedback is useful for system improvement, the terminal may send the feedback information to the server, including the analysis results of the sentiment engine.

[0285] Saving of Feedback and AI Retraining

[0286] The server saves the received feedback information in the database. The feedback information is used for retraining the generative AI model and the sentiment engine. The accuracy of responses to search requests and the accuracy of sentiment recognition improve for subsequent times. With this system, the crew can quickly and appropriately obtain the necessary manual information, not only improving the quality of customer service but also enabling flexible responses considering the user's emotional state.

[0287] As a specific example, when the keyword "Cash register operation new registration" is input, the server generates a manual including the operation procedure and returns information such as "To register a new product at the cash register, scan the product barcode, enter the price and quantity, and then click the registration button". By considering the user's emotional state throughout this process and reflecting it as appropriate feedback, the performance of the entire system is improved.

[0288] The flow of the specific process in Example 2 will be described using FIG. 13.

[0289] Step 1:

[0290] The server reads the connection settings and initializes the AI ​​model, emotion engine, and database connection. This process uses a Python-based Flask server, with TensorFlow or PyTorch used for the generated AI model, and OpenFace or EmotionAPI for the emotion engine.

[0291] Input: Connection configuration file

[0292] Output: System in initialization complete state

[0293] Specific actions: Read the server configuration file, initialize various libraries, and establish a database connection.

[0294] Step 2:

[0295] The user accesses the system and enters their username and password on the login screen. The terminal sends this authentication information to the server.

[0296] Input: Username, Password

[0297] Output: Session ID (if authentication is successful) or error message (if authentication fails)

[0298] Specific operation: The user enters information into the login form, and the device sends it to the server as a POST request. The server compares it with the database and generates a session ID.

[0299] Step 3:

[0300] The user enters keywords on the terminal and sends a search request. For example, they might enter "cash register operation new registration". The terminal converts these keywords into JSON format and sends it to the server.

[0301] Input: Keyword

[0302] Output: Search request (JSON format)

[0303] Specific operation: The user enters text in the search bar and presses the search button. The terminal uses the JSON.stringify function of JavaScript (registered trademark) to convert the keyword into JSON format and sends an AJAX request to the server.

[0304] Step 4:

[0305] The emotion engine analyzes the user's expression and voice at the time of input and recognizes the emotional state. The analysis result is transmitted to the server. For example, when the user inputs with a dissatisfied expression or voice, the emotion engine determines it as "unhappy".

[0306] Input: Expression data, voice data

[0307] Output: Emotional state (e.g., "unhappy", "joy", etc.)

[0308] Specific operation: The terminal's camera and microphone data are collected in real time, emotion analysis is performed using OpenCV and TensorFlow models, and the results are sent to the server in JSON format.

[0309] Step 5:

[0310] The server searches the database based on the received search request and emotion analysis result. If no relevant documents are found, a new manual content is generated using a generative AI.

[0311] Input: Search request, emotional state

[0312] Output: Formatted manual information

[0313] Specific operation: Executes SQL queries to search for relevant documents, and if necessary, sends prompts to generative AI models such as GPT-3 to generate new manual content. Adjusts tone and content based on emotional state.

[0314] Step 6:

[0315] The server sends formatted manual information to the terminal, which then displays it to the user.

[0316] Input: Formatted manual information

[0317] Output: Manual information displayed to the user

[0318] Specific operation: The server sends manual information to the terminal in JSON format, the terminal uses JavaScript to embed the data into HTML, and displays it to the user.

[0319] Step 7:

[0320] The user provides feedback on the information provided. For example, they might rate it as "This information was very helpful." The device sends this feedback, along with the sentiment engine's analysis results, to the server.

[0321] Input: Feedback, sentiment analysis results

[0322] Output: Sent feedback data

[0323] Specific operation: The user enters information into the feedback form and presses the submit button. The device sends the feedback data to the server in JSON format.

[0324] Step 8:

[0325] The server stores the received feedback information in a database. This feedback information is used to retrain generative AI models and emotion engines.

[0326] Input: Feedback data

[0327] Output: Saved feedback information, retrained AI model

[0328] Specific actions: Use SQL queries to save feedback data to the database, and then retrain the generative AI model and emotion engine using the new dataset.

[0329] (Application Example 2)

[0330] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0331] Conventional operational support systems in physical stores often fail to respond quickly and appropriately to problems faced by crew members, and a particular challenge is the lack of support that takes into account the emotional state of the crew. Furthermore, there are no systems that allow crew members to use advanced technologies such as voice input or smart glasses to support their work, which leads to decreased operational efficiency and negatively impacts the quality of customer service. This invention aims to improve operational efficiency and customer service quality by recognizing the emotional state of crew members and providing appropriate information.

[0332] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for reading connection settings and performing initialization, means for receiving authentication information from the user and performing authentication, means for receiving keywords from the user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, means for recognizing the user's emotional state using smart glasses, means for adjusting the tone of the generated information according to the user's emotional state, means for receiving questions from the user via voice input, and means for processing voice feedback. This enables the provision of appropriate information according to the emotional state of the crew and quick and effective work support using voice input.

[0333] "Connection settings" refer to the settings that include network configurations and authentication information necessary for each component of the system to communicate.

[0334] "Initialization" refers to the process of preparing the necessary settings and resources before a system starts operating, and ensuring that each component functions correctly.

[0335] User authentication is the process of verifying the authentication information (such as username and password) that a user enters to access a system, and confirming that they are a legitimate user.

[0336] A "keyword" refers to the words or phrases that users enter into the system as part of a search request.

[0337] A "search request" is a request made by a user to a system to retrieve specific information.

[0338] A "database" is a collection of data used by a system to manage and store information, enabling efficient information retrieval and updating.

[0339] "Related documents" refer to documents or guidelines that are directly related to what the user is looking for in their search request.

[0340] "Generative AI" is an artificial intelligence technology that automatically generates new information and documents based on user requests and emotional states.

[0341] "Formatting" refers to the process of adjusting the format and presentation of generated or retrieved information to make it easier for users to understand.

[0342] A "user terminal" refers to a device (such as a smartphone, tablet, or smart glasses) that a crew member uses to operate the system.

[0343] "Feedback" refers to the evaluations and comments that users give to the information provided or the function of the system.

[0344] "Retraining" is a learning process that uses collected feedback data to improve the performance of generative AI.

[0345] "Smart glasses" are wearable devices that users can wear and use to display and operate information visually and audibly.

[0346] "Emotional state" refers to the psychological state (e.g., joy, anger, fatigue, etc.) analyzed from the user's facial expressions and voice.

[0347] "Tone" refers to the tone of the language and expression used in the generated information, and it is adjusted according to the user's emotional state.

[0348] This invention is a system designed to support operations in retail stores, primarily using smart glasses to improve crew efficiency and customer service. The system consists of a server, user terminals (smart glasses), a generative AI model, an emotion engine, and a database.

[0349] Initialization and connection settings

[0350] The server loads connection settings and initializes the AI ​​model, emotion engine, and database connection when the system starts up. This process prepares the system so that each component can function correctly.

[0351] User Authentication

[0352] The user (crew member) wears smart glasses and authenticates by entering their username and password. The terminal sends this authentication information to the server, which then compares it with the database. If authentication is successful, the server generates a session ID and sends it back to the terminal.

[0353] Question input and sentiment recognition

[0354] When a user inputs a question by voice, the device sends the voice data to the server. Simultaneously, the smart glasses' camera and microphone are used to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's current emotional state.

[0355] Manual generation and information provision

[0356] The server uses a generative AI model to generate new information based on the received question and emotional state. It searches the database for relevant existing documents, and if no relevant documents are found, the generative AI model generates a new manual tailored to the user. The tone of the generated manual is adjusted to reflect the emotional state. The formatted information is displayed to the user through smart glasses.

[0357] Feedback gathering and retraining

[0358] The user provides voice feedback on the information provided. The device sends the feedback information to the server, which stores it in a database. The stored feedback information is used to retrain the generative AI model and emotion engine, improving the accuracy of future search requests and manual generation.

[0359] Hardware / software to use

[0360] Hardware:

[0361] Smart glasses (e.g., wearable devices)

[0362] server

[0363] software:

[0364] Flask (Web framework)

[0365] SQLite (database)

[0366] Emotion recognition engine

[0367] Generative AI models

[0368] Specific example

[0369] User authentication:

[0370] User: "Enter your username. Enter your password."

[0371] Question Input:

[0372] User: "Please tell me how to register a new product at the checkout."

[0373] Emotion recognition:

[0374] The emotion engine reads the user's tired facial expression and detects "fatigue."

[0375] Manual generation:

[0376] "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button. If you need more detailed instructions, please press the support button."

[0377] feedback:

[0378] User: "This information was very helpful."

[0379] Example of a prompt

[0380] "The user asked 'How do I register a new product at the register?' and the emotion 'fatigue' was detected. Please provide detailed instructions in an appropriate tone."

[0381] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0382] Step 1:

[0383] The server reads the connection settings and performs the necessary initialization. This prepares the entire system for normal operation. Specifically, this includes initializing the AI ​​model and emotion engine, and establishing the database connection. The input is the configuration file, and the output is the system's initialized state.

[0384] Step 2:

[0385] The user enters their username and password using the smart glasses interface. The device sends this information to the server. The server compares the entered authentication information with its database, and if authentication is successful, generates a session ID and sends it back to the device. The input is the user authentication information, and the output is the session ID.

[0386] Step 3:

[0387] The user inputs a question by voice. The device sends the voice data to the server and simultaneously uses the smart glasses' camera and microphone to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's emotional state. The input is voice data and facial expression data, and the output is the emotional state. The specific operation includes analysis by the emotion engine.

[0388] Step 4:

[0389] The server generates new information using a generative AI model based on the received question content and sentiment state. First, it searches the database for any relevant existing documents. If no relevant documents are found, the generative AI model generates a new manual using a prompt. The input is the question content and sentiment state, and the output is a formatted manual. The specific operation includes document generation by the generative AI model.

[0390] Step 5:

[0391] The generated manual has its tone adjusted according to the user's emotional state. The server sends this formatted information to the user's terminal (smart glasses). The terminal displays it to the user. The input is the unformatted manual information, and the output is the formatted manual displayed to the user. Specific actions include adjusting the tone of the information according to the user's emotions.

[0392] Step 6:

[0393] The user provides voice feedback on the information provided. The device sends the feedback information to the server. The server stores this information in a database. The input is voice feedback, and the output is the feedback information stored in the database.

[0394] Step 7:

[0395] The server uses stored feedback information to retrain the generative AI model and sentiment engine. This process improves the accuracy of subsequent search requests and manual generation. The input is feedback data, and the output is the retrained model. Specific actions include retraining the AI ​​model.

[0396] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0397] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0398] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0399] [Second Embodiment]

[0400] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0401] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0402] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0403] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0404] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0405] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0406] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0407] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0408] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0409] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0410] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0411] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0412] This invention relates to a system that uses generative AI to provide manuals for in-store sales operations to crew members. This system consists of multiple elements that work in conjunction with each other. The main components of the system and their specific operations are described below.

[0413] Initialization and User Authentication

[0414] The server reads the necessary connection settings when the system starts up and initializes the AI ​​model and database connection. When a user accesses the system, the server receives the user's authentication information (username, password) from the terminal and verifies it against the database. For users who are successfully authenticated, a unique session ID is generated and returned to the terminal.

[0415] Manual search request

[0416] The user enters a question or keyword into an input field on the terminal. For example, they might enter the keyword "cash register operation new registration". The terminal then sends this keyword to the server in JSON format.

[0417] Searching and generating manuals

[0418] The server analyzes the received keywords and searches the database for relevant documents. For example, if there is a manual corresponding to the keywords "cash register operation new registration," it retrieves its contents. If no relevant documents are found, the server uses generative AI to generate new manual content. The generated content is then formatted for easier reading.

[0419] Information return and display

[0420] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button," might be displayed.

[0421] Gathering feedback

[0422] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server. The server stores the feedback in a database and uses it to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[0423] Specific example

[0424] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[0425] 2. Manual Search: The user enters "cash register operation new registration," and the terminal sends the keywords to the server. The server searches the database and either retrieves the document or generates new manual content using generative AI. The generated information is formatted and sent to the terminal for display to the user.

[0426] 3. Feedback Collection: Users input feedback such as "Helpful" regarding the displayed information, and the device sends this to the server. The server stores the feedback in a database and uses it to retrain the generative AI.

[0427] This system allows even inexperienced crew members to quickly and appropriately obtain necessary manual information, thereby improving the quality of customer service.

[0428] The following describes the processing flow.

[0429] Step 1:

[0430] When the system starts up, the server reads the configuration file and initializes the AI ​​model and database connection. This makes the system's internal resources available.

[0431] Step 2:

[0432] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[0433] Step 3:

[0434] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[0435] Step 4:

[0436] The user enters a question or keyword into the search field on the terminal. For example, they might enter "cash register operation new registration".

[0437] Step 5:

[0438] The terminal sends the keywords entered by the user to the server in JSON format. An example request is {"query": "Register Operation New Registration", "session_id": "xyz123"}.

[0439] Step 6:

[0440] The server analyzes the received search request and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[0441] Step 7:

[0442] If the server cannot find relevant documents, it invokes a generative AI to generate new information. The generated information is specific and in a format that answers the user's questions.

[0443] Step 8:

[0444] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response.

[0445] Step 9:

[0446] The terminal displays the received manual information on its screen. For example, it might show detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[0447] Step 10:

[0448] Users provide feedback on the displayed information. For example, they might rate it as "This information was helpful."

[0449] Step 11:

[0450] The device sends user feedback information to the server in JSON format. An example request is {"manual_id": "abc123", "feedback": "Helpful", "session_id": "xyz123"}.

[0451] Step 12:

[0452] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI.

[0453] Step 13:

[0454] The server uses accumulated feedback to retrain the AI ​​model to improve the performance of the generative AI. This process improves the accuracy of responses to subsequent search requests.

[0455] (Example 1)

[0456] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0457] In retail sales operations, there is a problem where crew members cannot quickly and appropriately obtain information due to outdated or incomplete manuals. Furthermore, there are issues where manuals cannot be found, or newly generated information is insufficient, leading to a decline in the quality of customer service. Additionally, there is a problem where the quality of generated information does not improve because feedback on the usefulness of the information provided is not effectively utilized.

[0458] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0459] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI model if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in the database, means for retraining the generative AI model using the received feedback, means for initializing the AI ​​model and database connection when the system starts up, means for the user to enter keywords in a search field and the terminal to send those keywords to the server, and means for formatting the generated manual information and displaying it to the user. As a result, crew members can quickly obtain the latest and most appropriate manual information, improving the quality of customer service. Furthermore, it becomes possible to continuously improve the quality of the information generated by utilizing feedback.

[0460] "Connection settings" refers to a configuration file that initializes the system and contains connection information for necessary external resources and services.

[0461] "Initialization" refers to the process of preparing the necessary settings and resources when a system starts operating.

[0462] "Authentication information" refers to information such as usernames and passwords used to identify and authenticate users.

[0463] A "keyword" refers to a phrase or term that a user enters into a search field to search for specific information or manuals.

[0464] A "search request" refers to a request that a user sends to obtain specific information.

[0465] A "database" refers to a system that systematically stores information and data, and allows for searching and retrieval as needed.

[0466] "Document" refers to a piece of paper containing specific information, such as a manual or guidelines.

[0467] A "generative AI model" refers to an artificial intelligence model that generates new text or documents based on input information.

[0468] "Formatting" refers to adjusting the format of generated information to make it easier to see and read.

[0469] A "user terminal" refers to a device used by a user to input information or to check results.

[0470] "Feedback" refers to users' evaluations and opinions on the information provided.

[0471] "Retraining" refers to the process of improving the performance of an existing generative AI model by having it retrained based on user feedback.

[0472] "Submitting keywords" refers to the act of sending search terms entered by the user to the server.

[0473] "Formatting and displaying manual information to the user" refers to the process of converting generated or acquired manual information into a visually easy-to-read format and displaying it on the user's device.

[0474] This invention relates to a system that uses a generative AI model to provide manuals for in-store sales operations to crew members. The specific processing details and embodiments of the system are described below.

[0475] Initialization and User Authentication

[0476] When the system starts up, the server reads the connection configuration file (e.g., config.yaml) and initializes the connections to the AI ​​model (e.g., GPT-4) and the database (e.g., MySQL). This process prepares the AI ​​model and database for the system.

[0477] When a user accesses the system, the terminal receives authentication information (username and password) and sends it to the server in JSON format. The server compares the authentication information with the user table in the database, and if authentication is successful, generates a unique session ID and sends it back to the terminal.

[0478] Manual search request

[0479] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. The terminal sends this keyword to the server in JSON format.

[0480] Searching and generating manuals

[0481] The server analyzes the received keywords and uses a language analysis library (e.g., NLTK) to evaluate their relevance. Next, it searches the database for relevant manual documents. If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. The generated information is then formatted visually using a Python formatting library.

[0482] Information return and display

[0483] The server sends formatted manual information to the terminal in JSON format. The terminal converts the received data into a viewable format and displays it to the user. For example, it might display specific instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[0484] Gathering feedback

[0485] Users evaluate whether the information provided was helpful and enter feedback. For example, a rating such as "This information was helpful." The device sends this feedback to the server in JSON format. The server stores the received feedback in a database. This feedback is used to retrain generative AI models. Retraining is done to improve the model's response accuracy so that responses to subsequent search requests are more accurate.

[0486] Specific example

[0487] When a user enters their username and password on the login screen, the terminal sends this information to the server. If authentication is successful, the server generates a session ID and sends it back to the terminal. Next, the user enters keywords such as "cash register operation new registration," and the terminal sends these to the server. The server searches the database and either retrieves relevant documents or generates a new manual using a generative AI model. Finally, the terminal displays the generated information to the user, who enters feedback and sends it back to the system. Based on this feedback, the generative AI model is retrained.

[0488] Example of a prompt

[0489] "Register a new user for register operations"

[0490] "Procedure for registering new products at the register"

[0491] In this way, users can obtain the necessary information quickly and accurately. This can improve the efficiency of in-store sales operations and the quality of customer service.

[0492] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0493] Step 1:

[0494] Initialization

[0495] The server reads a connection configuration file (e.g., config.yaml) when the system starts up. It receives the path to the configuration file as input and sets the connection string and model parameters. Based on this, it initializes the connection to the AI ​​model (e.g., GPT-4) and the database (e.g., MySQL). It generates a log confirming the successful connection as output.

[0496] Step 2:

[0497] Receiving user authentication information

[0498] The user accesses the login screen and enters their username and password. The terminal receives this input information in JSON format and sends it to the server. This passes the authentication information to the server. The input consists of the username and password, and the output is JSON data containing the authentication information sent to the server.

[0499] Step 3:

[0500] Verification of authentication information

[0501] The server compares the received authentication information with the user table in the database. It uses the username and password as input to query the database and retrieve matching records. As output, if authentication is successful, it generates a unique session ID and sends it back to the terminal. If authentication fails, it returns an error message.

[0502] Step 4:

[0503] Keyword Input

[0504] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. This issues a search request to the system. The keyword entered by the user is passed to the terminal as input. As output, the search query data in JSON format is sent from the terminal to the server.

[0505] Step 5:

[0506] Keyword analysis and database search

[0507] The server extracts keywords from the received JSON data and parses them using a language analysis library (e.g., NLTK). The input is keywords, which are parsed to generate conditions for searching related documents. The output is a generated database query, which is then sent to the database.

[0508] Step 6:

[0509] Generating new manuals using generative AI models

[0510] If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. Keyword-based prompts (e.g., "new product register registration procedure") are sent to the AI ​​model as input. The generated text information is returned as output.

[0511] Step 7:

[0512] Information formatting

[0513] The server formats the generated text information to make it visually readable. The input is the result generated by an AI model, and formatting adjustments are made using a Python formatting library. The output is the formatted manual information.

[0514] Step 8:

[0515] Sending manual information

[0516] The server sends the formatted manual information to the terminal in JSON format. The input is the formatted manual information, and the output is a data packet generated to send this information to the terminal.

[0517] Step 9:

[0518] Display of manual information

[0519] The terminal converts the received manual information into a displayable format and displays it to the user. The input is manual information in JSON format, which is then rendered as HTML and UI components. The output is a visually displayed manual provided to the user.

[0520] Step 10:

[0521] Feedback Input

[0522] Users provide feedback on the information provided. For example, they rate the information they found helpful. The input is the feedback information, and the output is this information sent from the terminal to the server.

[0523] Step 11:

[0524] Save feedback

[0525] The server saves the received feedback to the database. The input is the feedback information, and the output is a log that confirms the save was successful.

[0526] Step 12:

[0527] Retraining generative AI models

[0528] The server retrains the generative AI model based on the feedback. The feedback data is used as input, serving as the dataset for training the AI ​​model. The output is a new model with improved performance.

[0529] (Application Example 1)

[0530] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0531] In autonomous vehicles, a challenge exists in that drivers and passengers often struggle to obtain quick and accurate information regarding driving operations and system settings when manuals are required. This problem is particularly pronounced in emergencies or when using an autonomous vehicle for the first time. Current information provision methods make it difficult to provide appropriate manuals in real time, which could compromise safety and convenience.

[0532] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0533] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and transmitting it to a user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, and means for providing manuals regarding driving operations and system settings on information terminals of autonomous vehicles and personal mobile terminals. This enables drivers and passengers to quickly and accurately obtain information regarding the operation of autonomous vehicles.

[0534] "Connection settings" refer to the specific settings that enable each component of a system to communicate with one another.

[0535] "Initialization" refers to the process of setting up the state of each component when a system starts up, preparing it for normal operation.

[0536] "Authentication information" refers to data used to verify a user's identity, and typically includes a username and password.

[0537] "Authentication" refers to the process of verifying whether a user has legitimate access rights based on the authentication information provided.

[0538] A "keyword" refers to a word or phrase that a user enters to search for specific information.

[0539] A "search request" refers to a request sent by a user to search for specific information.

[0540] A "database" refers to a system that systematically stores information so that it can be searched and retrieved later.

[0541] A "document" refers to a document that describes specific information or procedures.

[0542] "Generative AI" refers to a type of artificial intelligence that has the ability to generate new information and documents.

[0543] A "manual" refers to a document that explains specific operating procedures or system settings.

[0544] "Formatting" refers to the process of converting generated information into a format that is easy to see and understand.

[0545] A "user terminal" refers to a device used by a user to input information or view results. Examples include smartphones and vehicle information terminals.

[0546] "Feedback" refers to opinions, evaluations, and input regarding the effectiveness of information provided by users.

[0547] "Retraining" refers to the process of improving the performance of an existing AI model by having it learn again using new data.

[0548] An "autonomous vehicle" refers to a vehicle that has the ability to drive automatically without requiring direct control by a driver.

[0549] An "information terminal" refers to a device installed inside an autonomous vehicle that allows the user to input and confirm information.

[0550] The system of this invention is designed to quickly provide manuals regarding driving operations and system settings to information terminals in autonomous vehicles and to the driver's personal mobile device. The detailed processing steps and hardware and software used to implement this system are described below.

[0551] 1. Initialization and User Authentication

[0552] The server reads the necessary connection settings when the system starts up and initializes the database connection and generative AI model. When a user accesses the system, the server receives user authentication information (username, password) from the terminal and authenticates the user by comparing it with the database. For users who successfully authenticate, a unique session ID is generated and returned to the terminal. The hardware required here is a server and a personal mobile device, and the software utilizes a database system (e.g., MongoDB or SQL) and a generative AI model.

[0553] 2. Manual search request

[0554] The user enters a question or keyword into an input field on the device. For example, they might enter the keyword "How to turn on autonomous driving mode." The device sends this keyword to the server in JSON format. The server analyzes the received keyword and searches its database for relevant documents. The software used here utilizes a web framework (e.g., Flask).

[0555] 3. Searching for and generating manuals

[0556] The server searches the database for relevant documents based on the analyzed keywords. If no relevant documents are found, the server uses a generative AI model to generate new manual content. The generated content is then formatted for readability. The specific AI library used in this process is Some AI Library.

[0557] 4. Information return and display

[0558] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel," are displayed. Information terminals and smartphones are used as hardware for this purpose.

[0559] 5. Gathering feedback and retraining

[0560] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server, which stores it in a database. Furthermore, this feedback is used to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[0561] Examples of specific cases and prompt statements

[0562] As a concrete example, imagine a driver who wants to know "how to turn on autonomous driving mode" and inputs "Please tell me how to turn on autonomous driving mode." The system responds to this input as follows:

[0563] To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel.

[0564] In this way, users can quickly and accurately obtain information regarding the operation of autonomous vehicles.

[0565] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0566] Step 1:

[0567] The server reads the connection settings when the system starts up and performs database connection and generative AI model initialization. Specifically, the server reads DB connection information from the configuration file and establishes a connection to the database. It also reads the generative AI model from the specified path and completes its initialization. The input is the configuration file, and the output is the status of successful database connection and AI model initialization.

[0568] Step 2:

[0569] The user enters authentication information (username, password) from the terminal, and the terminal sends this information to the server. The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal. The input is the user's authentication information, and the output is the session ID or the authentication error status. Specifically, the server compares the authentication information with the user information in the database and generates a session ID if the match is successful.

[0570] Step 3:

[0571] The user enters a question or keyword into an input field on the device, and the device sends this keyword to the server in JSON format. The input is the keyword entered by the user, and the output is a search request in JSON format. Specifically, the device acquires the user's input in real time, formats it into JSON format, and sends it to the server.

[0572] Step 4:

[0573] The server parses the received keywords and searches the database for relevant documents. The input is a search request in JSON format, and the output is a flag indicating whether relevant documents or search results are not found. Specifically, the server creates a database query based on the keywords and performs a database search.

[0574] Step 5:

[0575] If no relevant documents are found in the database, the server uses a generative AI model to generate new manual content. The generated content is formatted and converted into a readable format. The input consists of flags and keywords indicating that no search results were found, and the output is the generated manual content. Specifically, the server inputs keywords as prompts to the generative AI model and receives the text generated by the AI ​​model. It then formats it into the appropriate format.

[0576] Step 6:

[0577] The server sends formatted manual information to the terminal, which then displays it to the user. The input is the formatted manual content, and the output is the specific operating procedure displayed to the user. Specifically, the server sends the manual content to the terminal, which receives it and displays it visually to the user.

[0578] Step 7:

[0579] The user evaluates whether the provided information was helpful and enters feedback. The terminal sends this feedback to the server. The input is the user's feedback, and the output is the transmission of the feedback data to the server. Specifically, the terminal retrieves the user's rating, formats it in JSON format, and sends it to the server.

[0580] Step 8:

[0581] The server stores the received feedback in a database and uses it to retrain the generative AI model. The input is the feedback data, and the output is the feedback stored in the database and a new dataset used for retraining. Specifically, the server stores the feedback in the database and periodically retrieves and uses the feedback data to retrain the generative AI model.

[0582] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0583] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system operates through a process from login to user emotion recognition, manual generation, feedback collection, and AI retraining. The main components and specific operation of the system are described below.

[0584] Initialization and User Authentication

[0585] When the system starts, the server loads the connection settings and initializes the AI ​​model, emotion engine, and database connection. When a user accesses the system, the terminal enters a username and password. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[0586] Manual search request

[0587] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format.

[0588] Emotion recognition by an emotion engine

[0589] At the same time, the emotion engine analyzes the user's facial expressions and voice during input to recognize their emotional state. For example, if a user makes an unhappy facial expression or makes an unhappy voice during input, the emotion engine analyzes this and determines that they are "unhappy."

[0590] Searching and generating manuals

[0591] The server analyzes the received search request and searches the database for relevant documents. If relevant documents are found, their contents are retrieved. If no documents are found, a generative AI is used to generate new manual content. Based on the analysis results of the emotion engine, the tone and content of the generated manual are adjusted.

[0592] Information return and display

[0593] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific instructions such as "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button" are displayed. If the information is deemed inappropriate, additional support information or actions (e.g., "View more detailed guide" or "Contact support center") are suggested.

[0594] Gathering feedback

[0595] Users provide feedback on the information provided, for example, by rating it as "This information was very helpful." The device sends the feedback information to the server, which may include the results of the sentiment engine's analysis.

[0596] Saving feedback and retraining the AI

[0597] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI model and the emotion engine, improving the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[0598] Specific example

[0599] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[0600] 2. Manual Search and Emotion Recognition: The user enters "cash register operation new registration," and the terminal sends the keyword along with facial expressions and voice data to the server. The server searches the database and, if necessary, generates new manual content using generative AI. The generated information reflects the analysis results of the emotion engine and is formatted in an appropriate tone and style. The formatted information is then displayed to the user through the terminal.

[0601] 3. Feedback Collection and Retraining: Users provide feedback such as "This was very helpful" regarding the information, and the device sends this to the server along with the analysis results of the emotion engine. The server stores the feedback and retrains the generative AI and emotion engine.

[0602] This system allows crew members to quickly and appropriately obtain necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state.

[0603] The following describes the processing flow.

[0604] Step 1:

[0605] When the system starts up, the server loads the connection settings and initializes the connections for the AI ​​model, emotion engine, and database. This makes the system's internal resources available.

[0606] Step 2:

[0607] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[0608] Step 3:

[0609] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[0610] Step 4:

[0611] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format. An example request is {"query": "cash register operation new registration", "session_id": "xyz123"}.

[0612] Step 5:

[0613] Simultaneously, the emotion engine collects the user's facial expressions and voice during input and analyzes the user's emotional state. For example, if the user's facial expression indicates displeasure, the emotion engine recognizes this as "unpleasant" and sends the analysis result to the server.

[0614] Step 6:

[0615] The server analyzes the received search request and sentiment state, and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[0616] Step 7:

[0617] If the server cannot find the relevant document in the database, it invokes a generative AI to generate new manual content. The generated content is then adjusted to an appropriate tone and content based on the analysis results of the emotion engine. For example, if the user is in an unpleasant state, the information will be generated using gentler and more polite language.

[0618] Step 8:

[0619] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response. For example, the response may also include additional support information tailored to the user's emotional state.

[0620] Step 9:

[0621] The terminal displays the received manual information on the screen. For example, it displays detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button." If the information is deemed inappropriate, additional support information (e.g., a link to a detailed guide or a direct contact number for support) is suggested.

[0622] Step 10:

[0623] Users provide feedback on the displayed information. For example, they might rate it as "This information was very helpful."

[0624] Step 11:

[0625] The device sends user feedback information to the server in JSON format. This information may include the results of the emotion engine's analysis. An example request is {"manual_id": "abc123", "feedback": "Helpful", "emotion": "happy", "session_id": "xyz123"}.

[0626] Step 12:

[0627] The server stores the received feedback information in a database and uses it to retrain the emotion engine and generative AI models. This data is used to improve the analysis accuracy and response quality of the AI ​​and emotion engine.

[0628] Step 13:

[0629] The server retrains its generative AI model and emotion engine based on accumulated feedback and emotion data. This improves the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[0630] This process allows crew members to quickly and appropriately obtain necessary manual information, as well as respond flexibly to the user's emotional state, thereby improving operational efficiency and customer satisfaction.

[0631] (Example 2)

[0632] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0633] In modern retail operations, manuals are needed to enable crew members to quickly and accurately obtain information. However, existing systems struggle to respond flexibly, taking into account the emotional state of users, and search results can sometimes be inaccurate. Furthermore, AI is not being retrained based on feedback, resulting in insufficient system performance improvements. A system is needed to address these challenges.

[0634] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0635] In this invention, the server includes means for reading and initializing connection settings; means for receiving and authenticating authentication information from the user; means for receiving keywords from the user and processing search requests; means for analyzing the user's facial expressions and voice during input and recognizing their emotional state; means for searching for relevant documents from a database; means for generating new information using generative AI if no relevant documents are found in the database; means for formatting the generated information based on the user's emotional state and sending it to the user terminal; means for receiving feedback from the user and storing it in a database; and means for retraining the generative AI using the received feedback. As a result, crew members can quickly and appropriately obtain the necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state. Furthermore, the accuracy of the AI ​​model can be continuously improved by utilizing the feedback.

[0636] "Connection settings" refer to the settings that include network and database connection information necessary for servers and terminals to communicate.

[0637] "Initialization" refers to the process of loading the necessary settings and data when a system starts up, and preparing the system to be operational.

[0638] "Authentication information" refers to usernames, passwords, and other authentication methods used to identify a user and grant them access.

[0639] A "search request" is a set of keywords or queries that a user sends to a server to retrieve specific information.

[0640] "Emotional state" refers to the state of emotions analyzed from the user's facial expressions and voice, and means feelings such as discomfort or joy.

[0641] "Generative AI" refers to artificial intelligence technology that has the ability to generate new information based on large amounts of data. Examples include natural language generation models.

[0642] A "user terminal" refers to a device, such as a computer or smartphone, that a user uses to access and operate a system.

[0643] "Feedback" refers to evaluations and opinions provided by users of a system, and is used to improve the system and enhance its performance.

[0644] "Retraining" is the process of using newly acquired data to train an artificial intelligence model again and improve its performance.

[0645] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system includes a series of processes from user login to emotion recognition, manual generation, feedback collection, and AI retraining.

[0646] Initialization and User Authentication

[0647] The server reads the connection settings when the system starts and initializes the AI ​​model, emotion engine, and database connection. The specific software used is a Python-based Flask server, and the generated AI model uses either TensorFlow or PyTorch. The emotion engine uses OpenFace or EmotionAPI. The user accesses the system and enters a username and password on the login screen. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[0648] Manual search request

[0649] The user enters a keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal converts this keyword into JSON format and sends it to the server.

[0650] Emotion recognition by an emotion engine

[0651] During this process, the emotion engine analyzes the user's facial expressions and voice. For example, if the user makes an unhappy facial expression or makes an unspoken voice, the emotion engine will determine this to be "unpleasant." OpenCV and TensorFlow models are used to collect and analyze the device's camera and microphone data in real time for emotion analysis.

[0652] Searching and generating manuals

[0653] The server analyzes the received search request and searches the database for relevant documents. If no matching documents are found in the database, it uses a generative AI (e.g., GPT-3) to generate new manual content. Based on the sentiment engine's analysis, it adjusts the tone and content of the generated manual. For example, if dissatisfaction is identified, it will include more detailed explanations and additional support information.

[0654] Information return and display

[0655] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific instructions such as "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button" are displayed. If the user deems the instructions inappropriate, additional support information or actions (e.g., "View more detailed guide" or "Contact support center") are suggested.

[0656] Gathering feedback

[0657] Users provide feedback on the information provided, for example, by rating it as "This information was very helpful." Because feedback helps improve the system, the device sends the feedback information to the server, which may include the results of the sentiment engine's analysis.

[0658] Saving feedback and retraining the AI

[0659] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI model and the emotion engine. This improves the accuracy of responses to subsequent search requests and the accuracy of emotion recognition. This system allows crew members to quickly and appropriately obtain necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state.

[0660] For example, if the keyword "register operation new registration" is entered, the server will generate a manual including the operation procedure and return information such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button." This entire process takes into account the user's emotional state and reflects it as appropriate feedback, thereby improving the overall performance of the system.

[0661] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0662] Step 1:

[0663] The server reads the connection settings and initializes the AI ​​model, emotion engine, and database connection. This process uses a Python-based Flask server, with TensorFlow or PyTorch used for the generated AI model, and OpenFace or EmotionAPI for the emotion engine.

[0664] Input: Connection configuration file

[0665] Output: System in initialization complete state

[0666] Specific actions: Read the server configuration file, initialize various libraries, and establish a database connection.

[0667] Step 2:

[0668] The user accesses the system and enters their username and password on the login screen. The terminal sends this authentication information to the server.

[0669] Input: Username, Password

[0670] Output: Session ID (if authentication is successful) or error message (if authentication fails)

[0671] Specific operation: The user enters information into the login form, and the device sends it to the server as a POST request. The server compares it with the database and generates a session ID.

[0672] Step 3:

[0673] The user enters keywords on the terminal and sends a search request. For example, they might enter "cash register operation new registration". The terminal converts these keywords into JSON format and sends it to the server.

[0674] Input: Keyword

[0675] Output: Search request (JSON format)

[0676] Specific operation: The user enters text into the search bar and presses the search button. The device uses the JavaScript JSON.stringify function to convert the keywords into JSON format and sends an AJAX request to the server.

[0677] Step 4:

[0678] The emotion engine analyzes the user's facial expressions and voice during input to recognize their emotional state. The analysis results are transmitted to the server. For example, if a user inputs with an unhappy facial expression or voice, the emotion engine will determine that they are "unhappy."

[0679] Input: Facial expression data, audio data

[0680] Output: Emotional state (e.g., "discomfort," "joy," etc.)

[0681] Specific operation: The device collects camera and microphone data in real time, performs sentiment analysis using OpenCV and TensorFlow models, and sends the results to the server in JSON format.

[0682] Step 5:

[0683] The server searches the database based on the received search request and sentiment analysis results. If no relevant documents are found, it uses a generative AI to generate new manual content.

[0684] Input: Search request, emotional state

[0685] Output: Formatted manual information

[0686] Specific operation: Executes SQL queries to search for relevant documents, and if necessary, sends prompts to generative AI models such as GPT-3 to generate new manual content. Adjusts tone and content based on emotional state.

[0687] Step 6:

[0688] The server sends formatted manual information to the terminal, which then displays it to the user.

[0689] Input: Formatted manual information

[0690] Output: Manual information displayed to the user

[0691] Specific operation: The server sends manual information to the terminal in JSON format, the terminal uses JavaScript to embed the data into HTML, and displays it to the user.

[0692] Step 7:

[0693] The user provides feedback on the information provided. For example, they might rate it as "This information was very helpful." The device sends this feedback, along with the sentiment engine's analysis results, to the server.

[0694] Input: Feedback, sentiment analysis results

[0695] Output: Sent feedback data

[0696] Specific operation: The user enters information into the feedback form and presses the submit button. The device sends the feedback data to the server in JSON format.

[0697] Step 8:

[0698] The server stores the received feedback information in a database. This feedback information is used to retrain generative AI models and emotion engines.

[0699] Input: Feedback data

[0700] Output: Saved feedback information, retrained AI model

[0701] Specific actions: Use SQL queries to save feedback data to the database, and then retrain the generative AI model and emotion engine using the new dataset.

[0702] (Application Example 2)

[0703] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0704] Conventional operational support systems in physical stores often fail to respond quickly and appropriately to problems faced by crew members, and a particular challenge is the lack of support that takes into account the emotional state of the crew. Furthermore, there are no systems that allow crew members to use advanced technologies such as voice input or smart glasses to support their work, which leads to decreased operational efficiency and negatively impacts the quality of customer service. This invention aims to improve operational efficiency and customer service quality by recognizing the emotional state of crew members and providing appropriate information.

[0705] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for reading connection settings and performing initialization, means for receiving authentication information from the user and performing authentication, means for receiving keywords from the user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, means for recognizing the user's emotional state using smart glasses, means for adjusting the tone of the generated information according to the user's emotional state, means for receiving questions from the user via voice input, and means for processing voice feedback. This enables the provision of appropriate information according to the emotional state of the crew and quick and effective work support using voice input.

[0706] "Connection settings" refer to the settings that include network configurations and authentication information necessary for each component of the system to communicate.

[0707] "Initialization" refers to the process of preparing the necessary settings and resources before a system starts operating, and ensuring that each component functions correctly.

[0708] User authentication is the process of verifying the authentication information (such as username and password) that a user enters to access a system, and confirming that they are a legitimate user.

[0709] A "keyword" refers to the words or phrases that users enter into the system as part of a search request.

[0710] A "search request" is a request made by a user to a system to retrieve specific information.

[0711] A "database" is a collection of data used by a system to manage and store information, enabling efficient information retrieval and updating.

[0712] "Related documents" refer to documents or guidelines that are directly related to what the user is looking for in their search request.

[0713] "Generative AI" is an artificial intelligence technology that automatically generates new information and documents based on user requests and emotional states.

[0714] "Formatting" refers to the process of adjusting the format and presentation of generated or retrieved information to make it easier for users to understand.

[0715] A "user terminal" refers to a device (such as a smartphone, tablet, or smart glasses) that a crew member uses to operate the system.

[0716] "Feedback" refers to the evaluations and comments that users give to the information provided or the function of the system.

[0717] "Retraining" is a learning process that uses collected feedback data to improve the performance of generative AI.

[0718] "Smart glasses" are wearable devices that users can wear and use to display and operate information visually and audibly.

[0719] "Emotional state" refers to the psychological state (e.g., joy, anger, fatigue, etc.) analyzed from the user's facial expressions and voice.

[0720] "Tone" refers to the tone of the language and expression used in the generated information, and it is adjusted according to the user's emotional state.

[0721] This invention is a system designed to support operations in retail stores, primarily using smart glasses to improve crew efficiency and customer service. The system consists of a server, user terminals (smart glasses), a generative AI model, an emotion engine, and a database.

[0722] Initialization and connection settings

[0723] The server loads connection settings and initializes the AI ​​model, emotion engine, and database connection when the system starts up. This process prepares the system so that each component can function correctly.

[0724] User Authentication

[0725] The user (crew member) wears smart glasses and authenticates by entering their username and password. The terminal sends this authentication information to the server, which then compares it with the database. If authentication is successful, the server generates a session ID and sends it back to the terminal.

[0726] Question input and sentiment recognition

[0727] When a user inputs a question by voice, the device sends the voice data to the server. Simultaneously, the smart glasses' camera and microphone are used to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's current emotional state.

[0728] Manual generation and information provision

[0729] The server uses a generative AI model to generate new information based on the received question and emotional state. It searches the database for relevant existing documents, and if no relevant documents are found, the generative AI model generates a new manual tailored to the user. The tone of the generated manual is adjusted to reflect the emotional state. The formatted information is displayed to the user through smart glasses.

[0730] Feedback gathering and retraining

[0731] The user provides voice feedback on the information provided. The device sends the feedback information to the server, which stores it in a database. The stored feedback information is used to retrain the generative AI model and emotion engine, improving the accuracy of future search requests and manual generation.

[0732] Hardware / software to use

[0733] Hardware:

[0734] Smart glasses (e.g., wearable devices)

[0735] server

[0736] software:

[0737] Flask (Web framework)

[0738] SQLite (database)

[0739] Emotion recognition engine

[0740] Generative AI models

[0741] Specific example

[0742] User authentication:

[0743] User: "Enter your username. Enter your password."

[0744] Question Input:

[0745] User: "Please tell me how to register a new product at the checkout."

[0746] Emotion recognition:

[0747] The emotion engine reads the user's tired facial expression and detects "fatigue."

[0748] Manual generation:

[0749] "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button. If you need more detailed instructions, please press the support button."

[0750] feedback:

[0751] User: "This information was very helpful."

[0752] Example of a prompt

[0753] "The user asked 'How do I register a new product at the register?' and the emotion 'fatigue' was detected. Please provide detailed instructions in an appropriate tone."

[0754] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0755] Step 1:

[0756] The server reads the connection settings and performs the necessary initialization. This prepares the entire system for normal operation. Specifically, this includes initializing the AI ​​model and emotion engine, and establishing the database connection. The input is the configuration file, and the output is the system's initialized state.

[0757] Step 2:

[0758] The user enters their username and password using the smart glasses interface. The device sends this information to the server. The server compares the entered authentication information with its database, and if authentication is successful, generates a session ID and sends it back to the device. The input is the user authentication information, and the output is the session ID.

[0759] Step 3:

[0760] The user inputs a question by voice. The device sends the voice data to the server and simultaneously uses the smart glasses' camera and microphone to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's emotional state. The input is voice data and facial expression data, and the output is the emotional state. The specific operation includes analysis by the emotion engine.

[0761] Step 4:

[0762] The server generates new information using a generative AI model based on the received question content and sentiment state. First, it searches the database for any relevant existing documents. If no relevant documents are found, the generative AI model generates a new manual using a prompt. The input is the question content and sentiment state, and the output is a formatted manual. The specific operation includes document generation by the generative AI model.

[0763] Step 5:

[0764] The generated manual has its tone adjusted according to the user's emotional state. The server sends this formatted information to the user's terminal (smart glasses). The terminal displays it to the user. The input is the unformatted manual information, and the output is the formatted manual displayed to the user. Specific actions include adjusting the tone of the information according to the user's emotions.

[0765] Step 6:

[0766] The user provides voice feedback on the information provided. The device sends the feedback information to the server. The server stores this information in a database. The input is voice feedback, and the output is the feedback information stored in the database.

[0767] Step 7:

[0768] The server uses stored feedback information to retrain the generative AI model and sentiment engine. This process improves the accuracy of subsequent search requests and manual generation. The input is feedback data, and the output is the retrained model. Specific actions include retraining the AI ​​model.

[0769] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0770] The data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One 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">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0771] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0772] [Third Embodiment]

[0773] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0774] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0775] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0776] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0777] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0778] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0779] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0780] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0781] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0782] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0783] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0784] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0785] This invention relates to a system that uses generative AI to provide manuals for in-store sales operations to crew members. This system consists of multiple elements that work in conjunction with each other. The main components of the system and their specific operations are described below.

[0786] Initialization and User Authentication

[0787] The server reads the necessary connection settings when the system starts up and initializes the AI ​​model and database connection. When a user accesses the system, the server receives the user's authentication information (username, password) from the terminal and verifies it against the database. For users who are successfully authenticated, a unique session ID is generated and returned to the terminal.

[0788] Manual search request

[0789] The user enters a question or keyword into an input field on the terminal. For example, they might enter the keyword "cash register operation new registration". The terminal then sends this keyword to the server in JSON format.

[0790] Searching and generating manuals

[0791] The server analyzes the received keywords and searches the database for relevant documents. For example, if there is a manual corresponding to the keywords "cash register operation new registration," it retrieves its contents. If no relevant documents are found, the server uses generative AI to generate new manual content. The generated content is then formatted for easier reading.

[0792] Information return and display

[0793] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button," might be displayed.

[0794] Gathering feedback

[0795] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server. The server stores the feedback in a database and uses it to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[0796] Specific example

[0797] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[0798] 2. Manual Search: The user enters "cash register operation new registration," and the terminal sends the keywords to the server. The server searches the database and either retrieves the document or generates new manual content using generative AI. The generated information is formatted and sent to the terminal for display to the user.

[0799] 3. Feedback Collection: Users input feedback such as "Helpful" regarding the displayed information, and the device sends this to the server. The server stores the feedback in a database and uses it to retrain the generative AI.

[0800] This system allows even inexperienced crew members to quickly and appropriately obtain necessary manual information, thereby improving the quality of customer service.

[0801] The following describes the processing flow.

[0802] Step 1:

[0803] When the system starts up, the server reads the configuration file and initializes the AI ​​model and database connection. This makes the system's internal resources available.

[0804] Step 2:

[0805] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[0806] Step 3:

[0807] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[0808] Step 4:

[0809] The user enters a question or keyword into the search field on the terminal. For example, they might enter "cash register operation new registration".

[0810] Step 5:

[0811] The terminal sends the keywords entered by the user to the server in JSON format. An example request is {"query": "Register Operation New Registration", "session_id": "xyz123"}.

[0812] Step 6:

[0813] The server analyzes the received search request and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[0814] Step 7:

[0815] If the server cannot find relevant documents, it invokes a generative AI to generate new information. The generated information is specific and in a format that answers the user's questions.

[0816] Step 8:

[0817] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response.

[0818] Step 9:

[0819] The terminal displays the received manual information on its screen. For example, it might show detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[0820] Step 10:

[0821] Users provide feedback on the displayed information. For example, they might rate it as "This information was helpful."

[0822] Step 11:

[0823] The device sends user feedback information to the server in JSON format. An example request is {"manual_id": "abc123", "feedback": "Helpful", "session_id": "xyz123"}.

[0824] Step 12:

[0825] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI.

[0826] Step 13:

[0827] The server uses accumulated feedback to retrain the AI ​​model to improve the performance of the generative AI. This process improves the accuracy of responses to subsequent search requests.

[0828] (Example 1)

[0829] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0830] In retail sales operations, there is a problem where crew members cannot quickly and appropriately obtain information due to outdated or incomplete manuals. Furthermore, there are issues where manuals cannot be found, or newly generated information is insufficient, leading to a decline in the quality of customer service. Additionally, there is a problem where the quality of generated information does not improve because feedback on the usefulness of the information provided is not effectively utilized.

[0831] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0832] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI model if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in the database, means for retraining the generative AI model using the received feedback, means for initializing the AI ​​model and database connection when the system starts up, means for the user to enter keywords in a search field and the terminal to send those keywords to the server, and means for formatting the generated manual information and displaying it to the user. As a result, crew members can quickly obtain the latest and most appropriate manual information, improving the quality of customer service. Furthermore, it becomes possible to continuously improve the quality of the information generated by utilizing feedback.

[0833] "Connection settings" refers to a configuration file that initializes the system and contains connection information for necessary external resources and services.

[0834] "Initialization" refers to the process of preparing the necessary settings and resources when a system starts operating.

[0835] "Authentication information" refers to information such as usernames and passwords used to identify and authenticate users.

[0836] A "keyword" refers to a phrase or term that a user enters into a search field to search for specific information or manuals.

[0837] A "search request" refers to a request that a user sends to obtain specific information.

[0838] A "database" refers to a system that systematically stores information and data, and allows for searching and retrieval as needed.

[0839] "Document" refers to a piece of paper containing specific information, such as a manual or guidelines.

[0840] A "generative AI model" refers to an artificial intelligence model that generates new text or documents based on input information.

[0841] "Formatting" refers to adjusting the format of generated information to make it easier to see and read.

[0842] A "user terminal" refers to a device used by a user to input information or to check results.

[0843] "Feedback" refers to users' evaluations and opinions on the information provided.

[0844] "Retraining" refers to the process of improving the performance of an existing generative AI model by having it retrained based on user feedback.

[0845] "Submitting keywords" refers to the act of sending search terms entered by the user to the server.

[0846] "Formatting and displaying manual information to the user" refers to the process of converting generated or acquired manual information into a visually easy-to-read format and displaying it on the user's device.

[0847] This invention relates to a system that uses a generative AI model to provide manuals for in-store sales operations to crew members. The specific processing details and embodiments of the system are described below.

[0848] Initialization and User Authentication

[0849] When the system starts up, the server reads the connection configuration file (e.g., config.yaml) and initializes the connections to the AI ​​model (e.g., GPT-4) and the database (e.g., MySQL). This process prepares the AI ​​model and database for the system.

[0850] When a user accesses the system, the terminal receives authentication information (username and password) and sends it to the server in JSON format. The server compares the authentication information with the user table in the database, and if authentication is successful, generates a unique session ID and sends it back to the terminal.

[0851] Manual search request

[0852] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. The terminal sends this keyword to the server in JSON format.

[0853] Searching and generating manuals

[0854] The server analyzes the received keywords and uses a language analysis library (e.g., NLTK) to evaluate their relevance. Next, it searches the database for relevant manual documents. If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. The generated information is then formatted visually using a Python formatting library.

[0855] Information return and display

[0856] The server sends formatted manual information to the terminal in JSON format. The terminal converts the received data into a viewable format and displays it to the user. For example, it might display specific instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[0857] Gathering feedback

[0858] Users evaluate whether the information provided was helpful and enter feedback. For example, a rating such as "This information was helpful." The device sends this feedback to the server in JSON format. The server stores the received feedback in a database. This feedback is used to retrain generative AI models. Retraining is done to improve the model's response accuracy so that responses to subsequent search requests are more accurate.

[0859] Specific example

[0860] When a user enters their username and password on the login screen, the terminal sends this information to the server. If authentication is successful, the server generates a session ID and sends it back to the terminal. Next, the user enters keywords such as "cash register operation new registration," and the terminal sends these to the server. The server searches the database and either retrieves relevant documents or generates a new manual using a generative AI model. Finally, the terminal displays the generated information to the user, who enters feedback and sends it back to the system. Based on this feedback, the generative AI model is retrained.

[0861] Example of a prompt

[0862] "Register a new user for register operations"

[0863] "Procedure for registering new products at the register"

[0864] In this way, users can obtain the necessary information quickly and accurately. This can improve the efficiency of in-store sales operations and the quality of customer service.

[0865] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0866] Step 1:

[0867] Initialization

[0868] The server reads a connection configuration file (e.g., config.yaml) when the system starts up. It receives the path to the configuration file as input and sets the connection string and model parameters. Based on this, it initializes the connection to the AI ​​model (e.g., GPT-4) and the database (e.g., MySQL). It generates a log confirming the successful connection as output.

[0869] Step 2:

[0870] Receiving user authentication information

[0871] The user accesses the login screen and enters their username and password. The terminal receives this input information in JSON format and sends it to the server. This passes the authentication information to the server. The input consists of the username and password, and the output is JSON data containing the authentication information sent to the server.

[0872] Step 3:

[0873] Verification of authentication information

[0874] The server compares the received authentication information with the user table in the database. It uses the username and password as input to query the database and retrieve matching records. As output, if authentication is successful, it generates a unique session ID and sends it back to the terminal. If authentication fails, it returns an error message.

[0875] Step 4:

[0876] Keyword Input

[0877] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. This issues a search request to the system. The keyword entered by the user is passed to the terminal as input. As output, the search query data in JSON format is sent from the terminal to the server.

[0878] Step 5:

[0879] Keyword analysis and database search

[0880] The server extracts keywords from the received JSON data and parses them using a language analysis library (e.g., NLTK). The input is keywords, which are parsed to generate conditions for searching related documents. The output is a generated database query, which is then sent to the database.

[0881] Step 6:

[0882] Generating new manuals using generative AI models

[0883] If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. Keyword-based prompts (e.g., "new product register registration procedure") are sent to the AI ​​model as input. The generated text information is returned as output.

[0884] Step 7:

[0885] Information formatting

[0886] The server formats the generated text information to make it visually readable. The input is the result generated by an AI model, and formatting adjustments are made using a Python formatting library. The output is the formatted manual information.

[0887] Step 8:

[0888] Sending manual information

[0889] The server sends the formatted manual information to the terminal in JSON format. The input is the formatted manual information, and the output is a data packet generated to send this information to the terminal.

[0890] Step 9:

[0891] Display of manual information

[0892] The terminal converts the received manual information into a displayable format and displays it to the user. The input is manual information in JSON format, which is then rendered as HTML and UI components. The output is a visually displayed manual provided to the user.

[0893] Step 10:

[0894] Feedback Input

[0895] Users provide feedback on the information provided. For example, they rate the information they found helpful. The input is the feedback information, and the output is this information sent from the terminal to the server.

[0896] Step 11:

[0897] Save feedback

[0898] The server saves the received feedback to the database. The input is the feedback information, and the output is a log that confirms the save was successful.

[0899] Step 12:

[0900] Retraining generative AI models

[0901] The server retrains the generative AI model based on the feedback. The feedback data is used as input, serving as the dataset for training the AI ​​model. The output is a new model with improved performance.

[0902] (Application Example 1)

[0903] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0904] In autonomous vehicles, a challenge exists in that drivers and passengers often struggle to obtain quick and accurate information regarding driving operations and system settings when manuals are required. This problem is particularly pronounced in emergencies or when using an autonomous vehicle for the first time. Current information provision methods make it difficult to provide appropriate manuals in real time, which could compromise safety and convenience.

[0905] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0906] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and transmitting it to a user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, and means for providing manuals regarding driving operations and system settings on information terminals of autonomous vehicles and personal mobile terminals. This enables drivers and passengers to quickly and accurately obtain information regarding the operation of autonomous vehicles.

[0907] "Connection settings" refer to the specific settings that enable each component of a system to communicate with one another.

[0908] "Initialization" refers to the process of setting up the state of each component when a system starts up, preparing it for normal operation.

[0909] "Authentication information" refers to data used to verify a user's identity, and typically includes a username and password.

[0910] "Authentication" refers to the process of verifying whether a user has legitimate access rights based on the authentication information provided.

[0911] A "keyword" refers to a word or phrase that a user enters to search for specific information.

[0912] A "search request" refers to a request sent by a user to search for specific information.

[0913] A "database" refers to a system that systematically stores information so that it can be searched and retrieved later.

[0914] A "document" refers to a document that describes specific information or procedures.

[0915] "Generative AI" refers to a type of artificial intelligence that has the ability to generate new information and documents.

[0916] A "manual" refers to a document that explains specific operating procedures or system settings.

[0917] "Formatting" refers to the process of converting generated information into a format that is easy to see and understand.

[0918] A "user terminal" refers to a device used by a user to input information or view results. Examples include smartphones and vehicle information terminals.

[0919] "Feedback" refers to opinions, evaluations, and input regarding the effectiveness of information provided by users.

[0920] "Retraining" refers to the process of improving the performance of an existing AI model by having it learn again using new data.

[0921] An "autonomous vehicle" refers to a vehicle that has the ability to drive automatically without requiring direct control by a driver.

[0922] An "information terminal" refers to a device installed inside an autonomous vehicle that allows the user to input and confirm information.

[0923] The system of this invention is designed to quickly provide manuals regarding driving operations and system settings to information terminals in autonomous vehicles and to the driver's personal mobile device. The detailed processing steps and hardware and software used to implement this system are described below.

[0924] 1. Initialization and User Authentication

[0925] The server reads the necessary connection settings when the system starts up and initializes the database connection and generative AI model. When a user accesses the system, the server receives user authentication information (username, password) from the terminal and authenticates the user by comparing it with the database. For users who successfully authenticate, a unique session ID is generated and returned to the terminal. The hardware required here is a server and a personal mobile device, and the software utilizes a database system (e.g., MongoDB or SQL) and a generative AI model.

[0926] 2. Manual search request

[0927] The user enters a question or keyword into an input field on the device. For example, they might enter the keyword "How to turn on autonomous driving mode." The device sends this keyword to the server in JSON format. The server analyzes the received keyword and searches its database for relevant documents. The software used here utilizes a web framework (e.g., Flask).

[0928] 3. Searching for and generating manuals

[0929] The server searches the database for relevant documents based on the analyzed keywords. If no relevant documents are found, the server uses a generative AI model to generate new manual content. The generated content is then formatted for readability. The specific AI library used in this process is Some AI Library.

[0930] 4. Information return and display

[0931] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel," are displayed. Information terminals and smartphones are used as hardware for this purpose.

[0932] 5. Gathering feedback and retraining

[0933] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server, which stores it in a database. Furthermore, this feedback is used to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[0934] Examples of specific cases and prompt statements

[0935] As a concrete example, imagine a driver who wants to know "how to turn on autonomous driving mode" and inputs "Please tell me how to turn on autonomous driving mode." The system responds to this input as follows:

[0936] To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel.

[0937] In this way, users can quickly and accurately obtain information regarding the operation of autonomous vehicles.

[0938] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0939] Step 1:

[0940] The server reads the connection settings when the system starts up and performs database connection and generative AI model initialization. Specifically, the server reads DB connection information from the configuration file and establishes a connection to the database. It also reads the generative AI model from the specified path and completes its initialization. The input is the configuration file, and the output is the status of successful database connection and AI model initialization.

[0941] Step 2:

[0942] The user enters authentication information (username, password) from the terminal, and the terminal sends this information to the server. The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal. The input is the user's authentication information, and the output is the session ID or the authentication error status. Specifically, the server compares the authentication information with the user information in the database and generates a session ID if the match is successful.

[0943] Step 3:

[0944] The user enters a question or keyword into an input field on the device, and the device sends this keyword to the server in JSON format. The input is the keyword entered by the user, and the output is a search request in JSON format. Specifically, the device acquires the user's input in real time, formats it into JSON format, and sends it to the server.

[0945] Step 4:

[0946] The server parses the received keywords and searches the database for relevant documents. The input is a search request in JSON format, and the output is a flag indicating whether relevant documents or search results are not found. Specifically, the server creates a database query based on the keywords and performs a database search.

[0947] Step 5:

[0948] If no relevant documents are found in the database, the server uses a generative AI model to generate new manual content. The generated content is formatted and converted into a readable format. The input consists of flags and keywords indicating that no search results were found, and the output is the generated manual content. Specifically, the server inputs keywords as prompts to the generative AI model and receives the text generated by the AI ​​model. It then formats it into the appropriate format.

[0949] Step 6:

[0950] The server sends formatted manual information to the terminal, which then displays it to the user. The input is the formatted manual content, and the output is the specific operating procedure displayed to the user. Specifically, the server sends the manual content to the terminal, which receives it and displays it visually to the user.

[0951] Step 7:

[0952] The user evaluates whether the provided information was helpful and enters feedback. The terminal sends this feedback to the server. The input is the user's feedback, and the output is the transmission of the feedback data to the server. Specifically, the terminal retrieves the user's rating, formats it in JSON format, and sends it to the server.

[0953] Step 8:

[0954] The server stores the received feedback in a database and uses it to retrain the generative AI model. The input is the feedback data, and the output is the feedback stored in the database and a new dataset used for retraining. Specifically, the server stores the feedback in the database and periodically retrieves and uses the feedback data to retrain the generative AI model.

[0955] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0956] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system operates through a process from login to user emotion recognition, manual generation, feedback collection, and AI retraining. The main components and specific operation of the system are described below.

[0957] Initialization and User Authentication

[0958] When the system starts, the server loads the connection settings and initializes the AI ​​model, emotion engine, and database connection. When a user accesses the system, the terminal enters a username and password. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[0959] Manual search request

[0960] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format.

[0961] Emotion recognition by an emotion engine

[0962] At the same time, the emotion engine analyzes the user's facial expressions and voice during input to recognize their emotional state. For example, if a user makes an unhappy facial expression or makes an unhappy voice during input, the emotion engine analyzes this and determines that they are "unhappy."

[0963] Searching and generating manuals

[0964] The server analyzes the received search request and searches the database for relevant documents. If relevant documents are found, their contents are retrieved. If no documents are found, a generative AI is used to generate new manual content. Based on the analysis results of the emotion engine, the tone and content of the generated manual are adjusted.

[0965] Information return and display

[0966] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific instructions such as "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button" are displayed. If the information is deemed inappropriate, additional support information or actions (e.g., "View more detailed guide" or "Contact support center") are suggested.

[0967] Gathering feedback

[0968] Users provide feedback on the information provided, for example, by rating it as "This information was very helpful." The device sends the feedback information to the server, which may include the results of the sentiment engine's analysis.

[0969] Saving feedback and retraining the AI

[0970] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI model and the emotion engine, improving the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[0971] Specific example

[0972] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[0973] 2. Manual Search and Emotion Recognition: The user enters "cash register operation new registration," and the terminal sends the keyword along with facial expressions and voice data to the server. The server searches the database and, if necessary, generates new manual content using generative AI. The generated information reflects the analysis results of the emotion engine and is formatted in an appropriate tone and style. The formatted information is then displayed to the user through the terminal.

[0974] 3. Feedback Collection and Retraining: Users provide feedback such as "This was very helpful" regarding the information, and the device sends this to the server along with the analysis results of the emotion engine. The server stores the feedback and retrains the generative AI and emotion engine.

[0975] This system allows crew members to quickly and appropriately obtain necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state.

[0976] The following describes the processing flow.

[0977] Step 1:

[0978] When the system starts up, the server loads the connection settings and initializes the connections for the AI ​​model, emotion engine, and database. This makes the system's internal resources available.

[0979] Step 2:

[0980] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[0981] Step 3:

[0982] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[0983] Step 4:

[0984] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format. An example request is {"query": "cash register operation new registration", "session_id": "xyz123"}.

[0985] Step 5:

[0986] Simultaneously, the emotion engine collects the user's facial expressions and voice during input and analyzes the user's emotional state. For example, if the user's facial expression indicates displeasure, the emotion engine recognizes this as "unpleasant" and sends the analysis result to the server.

[0987] Step 6:

[0988] The server analyzes the received search request and sentiment state, and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[0989] Step 7:

[0990] If the server cannot find the relevant document in the database, it invokes a generative AI to generate new manual content. The generated content is then adjusted to an appropriate tone and content based on the analysis results of the emotion engine. For example, if the user is in an unpleasant state, the information will be generated using gentler and more polite language.

[0991] Step 8:

[0992] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response. For example, the response may also include additional support information tailored to the user's emotional state.

[0993] Step 9:

[0994] The terminal displays the received manual information on the screen. For example, it displays detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button." If the information is deemed inappropriate, additional support information (e.g., a link to a detailed guide or a direct contact number for support) is suggested.

[0995] Step 10:

[0996] Users provide feedback on the displayed information. For example, they might rate it as "This information was very helpful."

[0997] Step 11:

[0998] The device sends user feedback information to the server in JSON format. This information may include the results of the emotion engine's analysis. An example request is {"manual_id": "abc123", "feedback": "Helpful", "emotion": "happy", "session_id": "xyz123"}.

[0999] Step 12:

[1000] The server stores the received feedback information in a database and uses it to retrain the emotion engine and generative AI models. This data is used to improve the analysis accuracy and response quality of the AI ​​and emotion engine.

[1001] Step 13:

[1002] The server retrains its generative AI model and emotion engine based on accumulated feedback and emotion data. This improves the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[1003] This process allows crew members to quickly and appropriately obtain necessary manual information, as well as respond flexibly to the user's emotional state, thereby improving operational efficiency and customer satisfaction.

[1004] (Example 2)

[1005] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1006] In modern retail operations, manuals are needed to enable crew members to quickly and accurately obtain information. However, existing systems struggle to respond flexibly, taking into account the emotional state of users, and search results can sometimes be inaccurate. Furthermore, AI is not being retrained based on feedback, resulting in insufficient system performance improvements. A system is needed to address these challenges.

[1007] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1008] In this invention, the server includes means for reading and initializing connection settings; means for receiving and authenticating authentication information from the user; means for receiving keywords from the user and processing search requests; means for analyzing the user's facial expressions and voice during input and recognizing their emotional state; means for searching for relevant documents from a database; means for generating new information using generative AI if no relevant documents are found in the database; means for formatting the generated information based on the user's emotional state and sending it to the user terminal; means for receiving feedback from the user and storing it in a database; and means for retraining the generative AI using the received feedback. As a result, crew members can quickly and appropriately obtain the necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state. Furthermore, the accuracy of the AI ​​model can be continuously improved by utilizing the feedback.

[1009] "Connection settings" refer to the settings that include network and database connection information necessary for servers and terminals to communicate.

[1010] "Initialization" refers to the process of loading the necessary settings and data when a system starts up, and preparing the system to be operational.

[1011] "Authentication information" refers to usernames, passwords, and other authentication methods used to identify a user and grant them access.

[1012] A "search request" is a set of keywords or queries that a user sends to a server to retrieve specific information.

[1013] "Emotional state" refers to the state of emotions analyzed from the user's facial expressions and voice, and means feelings such as discomfort or joy.

[1014] "Generative AI" refers to artificial intelligence technology that has the ability to generate new information based on large amounts of data. Examples include natural language generation models.

[1015] A "user terminal" refers to a device, such as a computer or smartphone, that a user uses to access and operate a system.

[1016] "Feedback" refers to evaluations and opinions provided by users of a system, and is used to improve the system and enhance its performance.

[1017] "Retraining" is the process of using newly acquired data to train an artificial intelligence model again and improve its performance.

[1018] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system includes a series of processes from user login to emotion recognition, manual generation, feedback collection, and AI retraining.

[1019] Initialization and User Authentication

[1020] The server reads the connection settings when the system starts and initializes the AI ​​model, emotion engine, and database connection. The specific software used is a Python-based Flask server, and the generated AI model uses either TensorFlow or PyTorch. The emotion engine uses OpenFace or EmotionAPI. The user accesses the system and enters a username and password on the login screen. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[1021] Manual search request

[1022] The user enters a keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal converts this keyword into JSON format and sends it to the server.

[1023] Emotion recognition by an emotion engine

[1024] During this process, the emotion engine analyzes the user's facial expressions and voice. For example, if the user makes an unhappy facial expression or makes an unspoken voice, the emotion engine will determine this to be "unpleasant." OpenCV and TensorFlow models are used to collect and analyze the device's camera and microphone data in real time for emotion analysis.

[1025] Searching and generating manuals

[1026] The server analyzes the received search request and searches the database for relevant documents. If no matching documents are found in the database, it uses a generative AI (e.g., GPT-3) to generate new manual content. Based on the sentiment engine's analysis, it adjusts the tone and content of the generated manual. For example, if dissatisfaction is identified, it will include more detailed explanations and additional support information.

[1027] Information return and display

[1028] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific instructions such as "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button" are displayed. If the user deems the instructions inappropriate, additional support information or actions (e.g., "View more detailed guide" or "Contact support center") are suggested.

[1029] Gathering feedback

[1030] Users provide feedback on the information provided, for example, by rating it as "This information was very helpful." Because feedback helps improve the system, the device sends the feedback information to the server, which may include the results of the sentiment engine's analysis.

[1031] Saving feedback and retraining the AI

[1032] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI model and the emotion engine. This improves the accuracy of responses to subsequent search requests and the accuracy of emotion recognition. This system allows crew members to quickly and appropriately obtain necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state.

[1033] For example, if the keyword "register operation new registration" is entered, the server will generate a manual including the operation procedure and return information such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button." This entire process takes into account the user's emotional state and reflects it as appropriate feedback, thereby improving the overall performance of the system.

[1034] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1035] Step 1:

[1036] The server reads the connection settings and initializes the AI ​​model, emotion engine, and database connection. This process uses a Python-based Flask server, with TensorFlow or PyTorch used for the generated AI model, and OpenFace or EmotionAPI for the emotion engine.

[1037] Input: Connection configuration file

[1038] Output: System in initialization complete state

[1039] Specific actions: Read the server configuration file, initialize various libraries, and establish a database connection.

[1040] Step 2:

[1041] The user accesses the system and enters their username and password on the login screen. The terminal sends this authentication information to the server.

[1042] Input: Username, Password

[1043] Output: Session ID (if authentication is successful) or error message (if authentication fails)

[1044] Specific operation: The user enters information into the login form, and the device sends it to the server as a POST request. The server compares it with the database and generates a session ID.

[1045] Step 3:

[1046] The user enters keywords on the terminal and sends a search request. For example, they might enter "cash register operation new registration". The terminal converts these keywords into JSON format and sends it to the server.

[1047] Input: Keyword

[1048] Output: Search request (JSON format)

[1049] Specific operation: The user enters text into the search bar and presses the search button. The device uses the JavaScript JSON.stringify function to convert the keywords into JSON format and sends an AJAX request to the server.

[1050] Step 4:

[1051] The emotion engine analyzes the user's facial expressions and voice during input to recognize their emotional state. The analysis results are transmitted to the server. For example, if a user inputs with an unhappy facial expression or voice, the emotion engine will determine that they are "unhappy."

[1052] Input: Facial expression data, audio data

[1053] Output: Emotional state (e.g., "discomfort," "joy," etc.)

[1054] Specific operation: The device collects camera and microphone data in real time, performs sentiment analysis using OpenCV and TensorFlow models, and sends the results to the server in JSON format.

[1055] Step 5:

[1056] The server searches the database based on the received search request and sentiment analysis results. If no relevant documents are found, it uses a generative AI to generate new manual content.

[1057] Input: Search request, emotional state

[1058] Output: Formatted manual information

[1059] Specific operation: Executes SQL queries to search for relevant documents, and if necessary, sends prompts to generative AI models such as GPT-3 to generate new manual content. Adjusts tone and content based on emotional state.

[1060] Step 6:

[1061] The server sends formatted manual information to the terminal, which then displays it to the user.

[1062] Input: Formatted manual information

[1063] Output: Manual information displayed to the user

[1064] Specific operation: The server sends manual information to the terminal in JSON format, the terminal uses JavaScript to embed the data into HTML, and displays it to the user.

[1065] Step 7:

[1066] The user provides feedback on the information provided. For example, they might rate it as "This information was very helpful." The device sends this feedback, along with the sentiment engine's analysis results, to the server.

[1067] Input: Feedback, sentiment analysis results

[1068] Output: Sent feedback data

[1069] Specific operation: The user enters information into the feedback form and presses the submit button. The device sends the feedback data to the server in JSON format.

[1070] Step 8:

[1071] The server stores the received feedback information in a database. This feedback information is used to retrain generative AI models and emotion engines.

[1072] Input: Feedback data

[1073] Output: Saved feedback information, retrained AI model

[1074] Specific actions: Use SQL queries to save feedback data to the database, and then retrain the generative AI model and emotion engine using the new dataset.

[1075] (Application Example 2)

[1076] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[1077] Conventional operational support systems in physical stores often fail to respond quickly and appropriately to problems faced by crew members, and a particular challenge is the lack of support that takes into account the emotional state of the crew. Furthermore, there are no systems that allow crew members to use advanced technologies such as voice input or smart glasses to support their work, which leads to decreased operational efficiency and negatively impacts the quality of customer service. This invention aims to improve operational efficiency and customer service quality by recognizing the emotional state of crew members and providing appropriate information.

[1078] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for reading connection settings and performing initialization, means for receiving authentication information from the user and performing authentication, means for receiving keywords from the user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, means for recognizing the user's emotional state using smart glasses, means for adjusting the tone of the generated information according to the user's emotional state, means for receiving questions from the user via voice input, and means for processing voice feedback. This enables the provision of appropriate information according to the emotional state of the crew and quick and effective work support using voice input.

[1079] "Connection settings" refer to the settings that include network configurations and authentication information necessary for each component of the system to communicate.

[1080] "Initialization" refers to the process of preparing the necessary settings and resources before a system starts operating, and ensuring that each component functions correctly.

[1081] User authentication is the process of verifying the authentication information (such as username and password) that a user enters to access a system, and confirming that they are a legitimate user.

[1082] A "keyword" refers to the words or phrases that users enter into the system as part of a search request.

[1083] A "search request" is a request made by a user to a system to retrieve specific information.

[1084] A "database" is a collection of data used by a system to manage and store information, enabling efficient information retrieval and updating.

[1085] "Related documents" refer to documents or guidelines that are directly related to what the user is looking for in their search request.

[1086] "Generative AI" is an artificial intelligence technology that automatically generates new information and documents based on user requests and emotional states.

[1087] "Formatting" refers to the process of adjusting the format and presentation of generated or retrieved information to make it easier for users to understand.

[1088] A "user terminal" refers to a device (such as a smartphone, tablet, or smart glasses) that a crew member uses to operate the system.

[1089] "Feedback" refers to the evaluations and comments that users give to the information provided or the function of the system.

[1090] "Retraining" is a learning process that uses collected feedback data to improve the performance of generative AI.

[1091] "Smart glasses" are wearable devices that users can wear and use to display and operate information visually and audibly.

[1092] "Emotional state" refers to the psychological state (e.g., joy, anger, fatigue, etc.) analyzed from the user's facial expressions and voice.

[1093] "Tone" refers to the tone of the language and expression used in the generated information, and it is adjusted according to the user's emotional state.

[1094] This invention is a system designed to support operations in retail stores, primarily using smart glasses to improve crew efficiency and customer service. The system consists of a server, user terminals (smart glasses), a generative AI model, an emotion engine, and a database.

[1095] Initialization and connection settings

[1096] The server loads connection settings and initializes the AI ​​model, emotion engine, and database connection when the system starts up. This process prepares the system so that each component can function correctly.

[1097] User Authentication

[1098] The user (crew member) wears smart glasses and authenticates by entering their username and password. The terminal sends this authentication information to the server, which then compares it with the database. If authentication is successful, the server generates a session ID and sends it back to the terminal.

[1099] Question input and sentiment recognition

[1100] When a user inputs a question by voice, the device sends the voice data to the server. Simultaneously, the smart glasses' camera and microphone are used to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's current emotional state.

[1101] Manual generation and information provision

[1102] The server uses a generative AI model to generate new information based on the received question and emotional state. It searches the database for relevant existing documents, and if no relevant documents are found, the generative AI model generates a new manual tailored to the user. The tone of the generated manual is adjusted to reflect the emotional state. The formatted information is displayed to the user through smart glasses.

[1103] Feedback gathering and retraining

[1104] The user provides voice feedback on the information provided. The device sends the feedback information to the server, which stores it in a database. The stored feedback information is used to retrain the generative AI model and emotion engine, improving the accuracy of future search requests and manual generation.

[1105] Hardware / software to use

[1106] Hardware:

[1107] Smart glasses (e.g., wearable devices)

[1108] server

[1109] software:

[1110] Flask (Web framework)

[1111] SQLite (database)

[1112] Emotion recognition engine

[1113] Generative AI models

[1114] Specific example

[1115] User authentication:

[1116] User: "Enter your username. Enter your password."

[1117] Question Input:

[1118] User: "Please tell me how to register a new product at the checkout."

[1119] Emotion recognition:

[1120] The emotion engine reads the user's tired facial expression and detects "fatigue."

[1121] Manual generation:

[1122] "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button. If you need more detailed instructions, please press the support button."

[1123] feedback:

[1124] User: "This information was very helpful."

[1125] Example of a prompt

[1126] "The user asked 'How do I register a new product at the register?' and the emotion 'fatigue' was detected. Please provide detailed instructions in an appropriate tone."

[1127] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1128] Step 1:

[1129] The server reads the connection settings and performs the necessary initialization. This prepares the entire system for normal operation. Specifically, this includes initializing the AI ​​model and emotion engine, and establishing the database connection. The input is the configuration file, and the output is the system's initialized state.

[1130] Step 2:

[1131] The user enters their username and password using the smart glasses interface. The device sends this information to the server. The server compares the entered authentication information with its database, and if authentication is successful, generates a session ID and sends it back to the device. The input is the user authentication information, and the output is the session ID.

[1132] Step 3:

[1133] The user inputs a question by voice. The device sends the voice data to the server and simultaneously uses the smart glasses' camera and microphone to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's emotional state. The input is voice data and facial expression data, and the output is the emotional state. The specific operation includes analysis by the emotion engine.

[1134] Step 4:

[1135] The server generates new information using a generative AI model based on the received question content and sentiment state. First, it searches the database for any relevant existing documents. If no relevant documents are found, the generative AI model generates a new manual using a prompt. The input is the question content and sentiment state, and the output is a formatted manual. The specific operation includes document generation by the generative AI model.

[1136] Step 5:

[1137] The generated manual has its tone adjusted according to the user's emotional state. The server sends this formatted information to the user's terminal (smart glasses). The terminal displays it to the user. The input is the unformatted manual information, and the output is the formatted manual displayed to the user. Specific actions include adjusting the tone of the information according to the user's emotions.

[1138] Step 6:

[1139] The user provides voice feedback on the information provided. The device sends the feedback information to the server. The server stores this information in a database. The input is voice feedback, and the output is the feedback information stored in the database.

[1140] Step 7:

[1141] The server uses stored feedback information to retrain the generative AI model and sentiment engine. This process improves the accuracy of subsequent search requests and manual generation. The input is feedback data, and the output is the retrained model. Specific actions include retraining the AI ​​model.

[1142] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1143] The data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One 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">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1144] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1145] [Fourth Embodiment]

[1146] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1147] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1148] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1149] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1150] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1151] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1152] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1153] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1154] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1155] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1156] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1157] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1158] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1159] This invention relates to a system that uses generative AI to provide manuals for in-store sales operations to crew members. This system consists of multiple elements that work in conjunction with each other. The main components of the system and their specific operations are described below.

[1160] Initialization and User Authentication

[1161] The server reads the necessary connection settings when the system starts up and initializes the AI ​​model and database connection. When a user accesses the system, the server receives the user's authentication information (username, password) from the terminal and verifies it against the database. For users who are successfully authenticated, a unique session ID is generated and returned to the terminal.

[1162] Manual search request

[1163] The user enters a question or keyword into an input field on the terminal. For example, they might enter the keyword "cash register operation new registration". The terminal then sends this keyword to the server in JSON format.

[1164] Searching and generating manuals

[1165] The server analyzes the received keywords and searches the database for relevant documents. For example, if there is a manual corresponding to the keywords "cash register operation new registration," it retrieves its contents. If no relevant documents are found, the server uses generative AI to generate new manual content. The generated content is then formatted for easier reading.

[1166] Information return and display

[1167] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button," might be displayed.

[1168] Gathering feedback

[1169] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server. The server stores the feedback in a database and uses it to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[1170] Specific example

[1171] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[1172] 2. Manual Search: The user enters "cash register operation new registration," and the terminal sends the keywords to the server. The server searches the database and either retrieves the document or generates new manual content using generative AI. The generated information is formatted and sent to the terminal for display to the user.

[1173] 3. Feedback Collection: Users input feedback such as "Helpful" regarding the displayed information, and the device sends this to the server. The server stores the feedback in a database and uses it to retrain the generative AI.

[1174] This system allows even inexperienced crew members to quickly and appropriately obtain necessary manual information, thereby improving the quality of customer service.

[1175] The following describes the processing flow.

[1176] Step 1:

[1177] When the system starts up, the server reads the configuration file and initializes the AI ​​model and database connection. This makes the system's internal resources available.

[1178] Step 2:

[1179] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[1180] Step 3:

[1181] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[1182] Step 4:

[1183] The user enters a question or keyword into the search field on the terminal. For example, they might enter "cash register operation new registration".

[1184] Step 5:

[1185] The terminal sends the keywords entered by the user to the server in JSON format. An example request is {"query": "Register Operation New Registration", "session_id": "xyz123"}.

[1186] Step 6:

[1187] The server analyzes the received search request and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[1188] Step 7:

[1189] If the server cannot find relevant documents, it invokes a generative AI to generate new information. The generated information is specific and in a format that answers the user's questions.

[1190] Step 8:

[1191] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response.

[1192] Step 9:

[1193] The terminal displays the received manual information on its screen. For example, it might show detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[1194] Step 10:

[1195] Users provide feedback on the displayed information. For example, they might rate it as "This information was helpful."

[1196] Step 11:

[1197] The device sends user feedback information to the server in JSON format. An example request is {"manual_id": "abc123", "feedback": "Helpful", "session_id": "xyz123"}.

[1198] Step 12:

[1199] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI.

[1200] Step 13:

[1201] The server uses accumulated feedback to retrain the AI ​​model to improve the performance of the generative AI. This process improves the accuracy of responses to subsequent search requests.

[1202] (Example 1)

[1203] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1204] In retail sales operations, there is a problem where crew members cannot quickly and appropriately obtain information due to outdated or incomplete manuals. Furthermore, there are issues where manuals cannot be found, or newly generated information is insufficient, leading to a decline in the quality of customer service. Additionally, there is a problem where the quality of generated information does not improve because feedback on the usefulness of the information provided is not effectively utilized.

[1205] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1206] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI model if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in the database, means for retraining the generative AI model using the received feedback, means for initializing the AI ​​model and database connection when the system starts up, means for the user to enter keywords in a search field and the terminal to send those keywords to the server, and means for formatting the generated manual information and displaying it to the user. As a result, crew members can quickly obtain the latest and most appropriate manual information, improving the quality of customer service. Furthermore, it becomes possible to continuously improve the quality of the information generated by utilizing feedback.

[1207] "Connection settings" refers to a configuration file that initializes the system and contains connection information for necessary external resources and services.

[1208] "Initialization" refers to the process of preparing the necessary settings and resources when a system starts operating.

[1209] "Authentication information" refers to information such as usernames and passwords used to identify and authenticate users.

[1210] A "keyword" refers to a phrase or term that a user enters into a search field to search for specific information or manuals.

[1211] A "search request" refers to a request that a user sends to obtain specific information.

[1212] A "database" refers to a system that systematically stores information and data, and allows for searching and retrieval as needed.

[1213] "Document" refers to a piece of paper containing specific information, such as a manual or guidelines.

[1214] A "generative AI model" refers to an artificial intelligence model that generates new text or documents based on input information.

[1215] "Formatting" refers to adjusting the format of generated information to make it easier to see and read.

[1216] A "user terminal" refers to a device used by a user to input information or to check results.

[1217] "Feedback" refers to users' evaluations and opinions on the information provided.

[1218] "Retraining" refers to the process of improving the performance of an existing generative AI model by having it retrained based on user feedback.

[1219] "Submitting keywords" refers to the act of sending search terms entered by the user to the server.

[1220] "Formatting and displaying manual information to the user" refers to the process of converting generated or acquired manual information into a visually easy-to-read format and displaying it on the user's device.

[1221] This invention relates to a system that uses a generative AI model to provide manuals for in-store sales operations to crew members. The specific processing details and embodiments of the system are described below.

[1222] Initialization and User Authentication

[1223] When the system starts up, the server reads the connection configuration file (e.g., config.yaml) and initializes the connections to the AI ​​model (e.g., GPT-4) and the database (e.g., MySQL). This process prepares the AI ​​model and database for the system.

[1224] When a user accesses the system, the terminal receives authentication information (username and password) and sends it to the server in JSON format. The server compares the authentication information with the user table in the database, and if authentication is successful, generates a unique session ID and sends it back to the terminal.

[1225] Manual search request

[1226] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. The terminal sends this keyword to the server in JSON format.

[1227] Searching and generating manuals

[1228] The server analyzes the received keywords and uses a language analysis library (e.g., NLTK) to evaluate their relevance. Next, it searches the database for relevant manual documents. If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. The generated information is then formatted visually using a Python formatting library.

[1229] Information return and display

[1230] The server sends formatted manual information to the terminal in JSON format. The terminal converts the received data into a viewable format and displays it to the user. For example, it might display specific instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button."

[1231] Gathering feedback

[1232] Users evaluate whether the information provided was helpful and enter feedback. For example, a rating such as "This information was helpful." The device sends this feedback to the server in JSON format. The server stores the received feedback in a database. This feedback is used to retrain generative AI models. Retraining is done to improve the model's response accuracy so that responses to subsequent search requests are more accurate.

[1233] Specific example

[1234] When a user enters their username and password on the login screen, the terminal sends this information to the server. If authentication is successful, the server generates a session ID and sends it back to the terminal. Next, the user enters keywords such as "cash register operation new registration," and the terminal sends these to the server. The server searches the database and either retrieves relevant documents or generates a new manual using a generative AI model. Finally, the terminal displays the generated information to the user, who enters feedback and sends it back to the system. Based on this feedback, the generative AI model is retrained.

[1235] Example of a prompt

[1236] "Register a new user for register operations"

[1237] "Procedure for registering new products at the register"

[1238] In this way, users can obtain the necessary information quickly and accurately. This can improve the efficiency of in-store sales operations and the quality of customer service.

[1239] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1240] Step 1:

[1241] Initialization

[1242] The server reads a connection configuration file (e.g., config.yaml) when the system starts up. It receives the path to the configuration file as input and sets the connection string and model parameters. Based on this, it initializes the connection to the AI ​​model (e.g., GPT-4) and the database (e.g., MySQL). It generates a log confirming the successful connection as output.

[1243] Step 2:

[1244] Receiving user authentication information

[1245] The user accesses the login screen and enters their username and password. The terminal receives this input information in JSON format and sends it to the server. This passes the authentication information to the server. The input consists of the username and password, and the output is JSON data containing the authentication information sent to the server.

[1246] Step 3:

[1247] Verification of authentication information

[1248] The server compares the received authentication information with the user table in the database. It uses the username and password as input to query the database and retrieve matching records. As output, if authentication is successful, it generates a unique session ID and sends it back to the terminal. If authentication fails, it returns an error message.

[1249] Step 4:

[1250] Keyword Input

[1251] The user enters a keyword (e.g., "cash register operation new registration") into an input field on the terminal. This issues a search request to the system. The keyword entered by the user is passed to the terminal as input. As output, the search query data in JSON format is sent from the terminal to the server.

[1252] Step 5:

[1253] Keyword analysis and database search

[1254] The server extracts keywords from the received JSON data and parses them using a language analysis library (e.g., NLTK). The input is keywords, which are parsed to generate conditions for searching related documents. The output is a generated database query, which is then sent to the database.

[1255] Step 6:

[1256] Generating new manuals using generative AI models

[1257] If no relevant documents are found in the database, the server uses a generative AI model (e.g., GPT-4) to generate new manual information. Keyword-based prompts (e.g., "new product register registration procedure") are sent to the AI ​​model as input. The generated text information is returned as output.

[1258] Step 7:

[1259] Information formatting

[1260] The server formats the generated text information to make it visually readable. The input is the result generated by an AI model, and formatting adjustments are made using a Python formatting library. The output is the formatted manual information.

[1261] Step 8:

[1262] Sending manual information

[1263] The server sends the formatted manual information to the terminal in JSON format. The input is the formatted manual information, and the output is a data packet generated to send this information to the terminal.

[1264] Step 9:

[1265] Display of manual information

[1266] The terminal converts the received manual information into a displayable format and displays it to the user. The input is manual information in JSON format, which is then rendered as HTML and UI components. The output is a visually displayed manual provided to the user.

[1267] Step 10:

[1268] Feedback Input

[1269] Users provide feedback on the information provided. For example, they rate the information they found helpful. The input is the feedback information, and the output is this information sent from the terminal to the server.

[1270] Step 11:

[1271] Save feedback

[1272] The server saves the received feedback to the database. The input is the feedback information, and the output is a log that confirms the save was successful.

[1273] Step 12:

[1274] Retraining generative AI models

[1275] The server retrains the generative AI model based on the feedback. The feedback data is used as input, serving as the dataset for training the AI ​​model. The output is a new model with improved performance.

[1276] (Application Example 1)

[1277] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1278] In autonomous vehicles, a challenge exists in that drivers and passengers often struggle to obtain quick and accurate information regarding driving operations and system settings when manuals are required. This problem is particularly pronounced in emergencies or when using an autonomous vehicle for the first time. Current information provision methods make it difficult to provide appropriate manuals in real time, which could compromise safety and convenience.

[1279] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1280] In this invention, the server includes means for reading and initializing connection settings, means for receiving and authenticating authentication information from a user, means for receiving keywords from a user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and transmitting it to a user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, and means for providing manuals regarding driving operations and system settings on information terminals of autonomous vehicles and personal mobile terminals. This enables drivers and passengers to quickly and accurately obtain information regarding the operation of autonomous vehicles.

[1281] "Connection settings" refer to the specific settings that enable each component of a system to communicate with one another.

[1282] "Initialization" refers to the process of setting up the state of each component when a system starts up, preparing it for normal operation.

[1283] "Authentication information" refers to data used to verify a user's identity, and typically includes a username and password.

[1284] "Authentication" refers to the process of verifying whether a user has legitimate access rights based on the authentication information provided.

[1285] A "keyword" refers to a word or phrase that a user enters to search for specific information.

[1286] A "search request" refers to a request sent by a user to search for specific information.

[1287] A "database" refers to a system that systematically stores information so that it can be searched and retrieved later.

[1288] A "document" refers to a document that describes specific information or procedures.

[1289] "Generative AI" refers to a type of artificial intelligence that has the ability to generate new information and documents.

[1290] A "manual" refers to a document that explains specific operating procedures or system settings.

[1291] "Formatting" refers to the process of converting generated information into a format that is easy to see and understand.

[1292] A "user terminal" refers to a device used by a user to input information or view results. Examples include smartphones and vehicle information terminals.

[1293] "Feedback" refers to opinions, evaluations, and input regarding the effectiveness of information provided by users.

[1294] "Retraining" refers to the process of improving the performance of an existing AI model by having it learn again using new data.

[1295] An "autonomous vehicle" refers to a vehicle that has the ability to drive automatically without requiring direct control by a driver.

[1296] An "information terminal" refers to a device installed inside an autonomous vehicle that allows the user to input and confirm information.

[1297] The system of this invention is designed to quickly provide manuals regarding driving operations and system settings to information terminals in autonomous vehicles and to the driver's personal mobile device. The detailed processing steps and hardware and software used to implement this system are described below.

[1298] 1. Initialization and User Authentication

[1299] The server reads the necessary connection settings when the system starts up and initializes the database connection and generative AI model. When a user accesses the system, the server receives user authentication information (username, password) from the terminal and authenticates the user by comparing it with the database. For users who successfully authenticate, a unique session ID is generated and returned to the terminal. The hardware required here is a server and a personal mobile device, and the software utilizes a database system (e.g., MongoDB or SQL) and a generative AI model.

[1300] 2. Manual search request

[1301] The user enters a question or keyword into an input field on the device. For example, they might enter the keyword "How to turn on autonomous driving mode." The device sends this keyword to the server in JSON format. The server analyzes the received keyword and searches its database for relevant documents. The software used here utilizes a web framework (e.g., Flask).

[1302] 3. Searching for and generating manuals

[1303] The server searches the database for relevant documents based on the analyzed keywords. If no relevant documents are found, the server uses a generative AI model to generate new manual content. The generated content is then formatted for readability. The specific AI library used in this process is Some AI Library.

[1304] 4. Information return and display

[1305] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific operating instructions such as, "To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel," are displayed. Information terminals and smartphones are used as hardware for this purpose.

[1306] 5. Gathering feedback and retraining

[1307] Users evaluate whether the information provided was helpful and provide feedback. For example, they might rate it as "This information was helpful." The device sends this feedback to the server, which stores it in a database. Furthermore, this feedback is used to retrain the generative AI model. This process improves the accuracy of responses to subsequent search requests.

[1308] Examples of specific cases and prompt statements

[1309] As a concrete example, imagine a driver who wants to know "how to turn on autonomous driving mode" and inputs "Please tell me how to turn on autonomous driving mode." The system responds to this input as follows:

[1310] To turn on autonomous driving mode, tap the autonomous driving icon on the vehicle's information display and select 'Yes' in response to the confirmation message. Then, press the autonomous driving button on the steering wheel.

[1311] In this way, users can quickly and accurately obtain information regarding the operation of autonomous vehicles.

[1312] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1313] Step 1:

[1314] The server reads the connection settings when the system starts up and performs database connection and generative AI model initialization. Specifically, the server reads DB connection information from the configuration file and establishes a connection to the database. It also reads the generative AI model from the specified path and completes its initialization. The input is the configuration file, and the output is the status of successful database connection and AI model initialization.

[1315] Step 2:

[1316] The user enters authentication information (username, password) from the terminal, and the terminal sends this information to the server. The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal. The input is the user's authentication information, and the output is the session ID or the authentication error status. Specifically, the server compares the authentication information with the user information in the database and generates a session ID if the match is successful.

[1317] Step 3:

[1318] The user enters a question or keyword into an input field on the device, and the device sends this keyword to the server in JSON format. The input is the keyword entered by the user, and the output is a search request in JSON format. Specifically, the device acquires the user's input in real time, formats it into JSON format, and sends it to the server.

[1319] Step 4:

[1320] The server parses the received keywords and searches the database for relevant documents. The input is a search request in JSON format, and the output is a flag indicating whether relevant documents or search results are not found. Specifically, the server creates a database query based on the keywords and performs a database search.

[1321] Step 5:

[1322] If no relevant documents are found in the database, the server uses a generative AI model to generate new manual content. The generated content is formatted and converted into a readable format. The input consists of flags and keywords indicating that no search results were found, and the output is the generated manual content. Specifically, the server inputs keywords as prompts to the generative AI model and receives the text generated by the AI ​​model. It then formats it into the appropriate format.

[1323] Step 6:

[1324] The server sends formatted manual information to the terminal, which then displays it to the user. The input is the formatted manual content, and the output is the specific operating procedure displayed to the user. Specifically, the server sends the manual content to the terminal, which receives it and displays it visually to the user.

[1325] Step 7:

[1326] The user evaluates whether the provided information was helpful and enters feedback. The terminal sends this feedback to the server. The input is the user's feedback, and the output is the transmission of the feedback data to the server. Specifically, the terminal retrieves the user's rating, formats it in JSON format, and sends it to the server.

[1327] Step 8:

[1328] The server stores the received feedback in a database and uses it to retrain the generative AI model. The input is the feedback data, and the output is the feedback stored in the database and a new dataset used for retraining. Specifically, the server stores the feedback in the database and periodically retrieves and uses the feedback data to retrain the generative AI model.

[1329] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1330] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system operates through a process from login to user emotion recognition, manual generation, feedback collection, and AI retraining. The main components and specific operation of the system are described below.

[1331] Initialization and User Authentication

[1332] When the system starts, the server loads the connection settings and initializes the AI ​​model, emotion engine, and database connection. When a user accesses the system, the terminal enters a username and password. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[1333] Manual search request

[1334] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format.

[1335] Emotion recognition by an emotion engine

[1336] At the same time, the emotion engine analyzes the user's facial expressions and voice during input to recognize their emotional state. For example, if a user makes an unhappy facial expression or makes an unhappy voice during input, the emotion engine analyzes this and determines that they are "unhappy."

[1337] Searching and generating manuals

[1338] The server analyzes the received search request and searches the database for relevant documents. If relevant documents are found, their contents are retrieved. If no documents are found, a generative AI is used to generate new manual content. Based on the analysis results of the emotion engine, the tone and content of the generated manual are adjusted.

[1339] Information return and display

[1340] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific instructions such as "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button" are displayed. If the information is deemed inappropriate, additional support information or actions (e.g., "View more detailed guide" or "Contact support center") are suggested.

[1341] Gathering feedback

[1342] Users provide feedback on the information provided, for example, by rating it as "This information was very helpful." The device sends the feedback information to the server, which may include the results of the sentiment engine's analysis.

[1343] Saving feedback and retraining the AI

[1344] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI model and the emotion engine, improving the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[1345] Specific example

[1346] 1. User Authentication: The user enters their username and password on the login screen, and the device sends this information to the server. The server verifies the authentication information, generates a session ID, and returns it to the device.

[1347] 2. Manual Search and Emotion Recognition: The user enters "cash register operation new registration," and the terminal sends the keyword along with facial expressions and voice data to the server. The server searches the database and, if necessary, generates new manual content using generative AI. The generated information reflects the analysis results of the emotion engine and is formatted in an appropriate tone and style. The formatted information is then displayed to the user through the terminal.

[1348] 3. Feedback Collection and Retraining: Users provide feedback such as "This was very helpful" regarding the information, and the device sends this to the server along with the analysis results of the emotion engine. The server stores the feedback and retrains the generative AI and emotion engine.

[1349] This system allows crew members to quickly and appropriately obtain necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state.

[1350] The following describes the processing flow.

[1351] Step 1:

[1352] When the system starts up, the server loads the connection settings and initializes the connections for the AI ​​model, emotion engine, and database. This makes the system's internal resources available.

[1353] Step 2:

[1354] The user opens the login screen on their device and enters their username and password. The device then sends this authentication information to the server.

[1355] Step 3:

[1356] The server compares the received authentication information with the database and performs authentication. If authentication is successful, the server generates a unique session ID and returns it to the terminal.

[1357] Step 4:

[1358] The user enters a question or keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal sends this keyword to the server in JSON format. An example request is {"query": "cash register operation new registration", "session_id": "xyz123"}.

[1359] Step 5:

[1360] Simultaneously, the emotion engine collects the user's facial expressions and voice during input and analyzes the user's emotional state. For example, if the user's facial expression indicates displeasure, the emotion engine recognizes this as "unpleasant" and sends the analysis result to the server.

[1361] Step 6:

[1362] The server analyzes the received search request and sentiment state, and searches the database for relevant documents. For example, it might search for a manual corresponding to "cash register operation new registration".

[1363] Step 7:

[1364] If the server cannot find the relevant document in the database, it invokes a generative AI to generate new manual content. The generated content is then adjusted to an appropriate tone and content based on the analysis results of the emotion engine. For example, if the user is in an unpleasant state, the information will be generated using gentler and more polite language.

[1365] Step 8:

[1366] The server formats the retrieved or generated manual information and sends it to the terminal as a JSON response. For example, the response may also include additional support information tailored to the user's emotional state.

[1367] Step 9:

[1368] The terminal displays the received manual information on the screen. For example, it displays detailed instructions such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button." If the information is deemed inappropriate, additional support information (e.g., a link to a detailed guide or a direct contact number for support) is suggested.

[1369] Step 10:

[1370] Users provide feedback on the displayed information. For example, they might rate it as "This information was very helpful."

[1371] Step 11:

[1372] The device sends user feedback information to the server in JSON format. This information may include the results of the emotion engine's analysis. An example request is {"manual_id": "abc123", "feedback": "Helpful", "emotion": "happy", "session_id": "xyz123"}.

[1373] Step 12:

[1374] The server stores the received feedback information in a database and uses it to retrain the emotion engine and generative AI models. This data is used to improve the analysis accuracy and response quality of the AI ​​and emotion engine.

[1375] Step 13:

[1376] The server retrains its generative AI model and emotion engine based on accumulated feedback and emotion data. This improves the accuracy of responses to subsequent search requests and the accuracy of emotion recognition.

[1377] This process allows crew members to quickly and appropriately obtain necessary manual information, as well as respond flexibly to the user's emotional state, thereby improving operational efficiency and customer satisfaction.

[1378] (Example 2)

[1379] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1380] In modern retail operations, manuals are needed to enable crew members to quickly and accurately obtain information. However, existing systems struggle to respond flexibly, taking into account the emotional state of users, and search results can sometimes be inaccurate. Furthermore, AI is not being retrained based on feedback, resulting in insufficient system performance improvements. A system is needed to address these challenges.

[1381] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1382] In this invention, the server includes means for reading and initializing connection settings; means for receiving and authenticating authentication information from the user; means for receiving keywords from the user and processing search requests; means for analyzing the user's facial expressions and voice during input and recognizing their emotional state; means for searching for relevant documents from a database; means for generating new information using generative AI if no relevant documents are found in the database; means for formatting the generated information based on the user's emotional state and sending it to the user terminal; means for receiving feedback from the user and storing it in a database; and means for retraining the generative AI using the received feedback. As a result, crew members can quickly and appropriately obtain the necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state. Furthermore, the accuracy of the AI ​​model can be continuously improved by utilizing the feedback.

[1383] "Connection settings" refer to the settings that include network and database connection information necessary for servers and terminals to communicate.

[1384] "Initialization" refers to the process of loading the necessary settings and data when a system starts up, and preparing the system to be operational.

[1385] "Authentication information" refers to usernames, passwords, and other authentication methods used to identify a user and grant them access.

[1386] A "search request" is a set of keywords or queries that a user sends to a server to retrieve specific information.

[1387] "Emotional state" refers to the state of emotions analyzed from the user's facial expressions and voice, and means feelings such as discomfort or joy.

[1388] "Generative AI" refers to artificial intelligence technology that has the ability to generate new information based on large amounts of data. Examples include natural language generation models.

[1389] A "user terminal" refers to a device, such as a computer or smartphone, that a user uses to access and operate a system.

[1390] "Feedback" refers to evaluations and opinions provided by users of a system, and is used to improve the system and enhance its performance.

[1391] "Retraining" is the process of using newly acquired data to train an artificial intelligence model again and improve its performance.

[1392] This invention relates to a system that combines generative AI and an emotion engine to support the work of crew members in retail sales operations. This system includes a series of processes from user login to emotion recognition, manual generation, feedback collection, and AI retraining.

[1393] Initialization and User Authentication

[1394] The server reads the connection settings when the system starts and initializes the AI ​​model, emotion engine, and database connection. The specific software used is a Python-based Flask server, and the generated AI model uses either TensorFlow or PyTorch. The emotion engine uses OpenFace or EmotionAPI. The user accesses the system and enters a username and password on the login screen. The terminal sends this authentication information to the server, which verifies it against the database. If authentication is successful, a session ID is generated and returned to the terminal.

[1395] Manual search request

[1396] The user enters a keyword on the terminal. For example, they might enter "cash register operation new registration". The terminal converts this keyword into JSON format and sends it to the server.

[1397] Emotion recognition by an emotion engine

[1398] During this process, the emotion engine analyzes the user's facial expressions and voice. For example, if the user makes an unhappy facial expression or makes an unspoken voice, the emotion engine will determine this to be "unpleasant." OpenCV and TensorFlow models are used to collect and analyze the device's camera and microphone data in real time for emotion analysis.

[1399] Searching and generating manuals

[1400] The server analyzes the received search request and searches the database for relevant documents. If no matching documents are found in the database, it uses a generative AI (e.g., GPT-3) to generate new manual content. Based on the sentiment engine's analysis, it adjusts the tone and content of the generated manual. For example, if dissatisfaction is identified, it will include more detailed explanations and additional support information.

[1401] Information return and display

[1402] The formatted manual information is sent from the server to the terminal, which then displays it to the user. For example, specific instructions such as "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button" are displayed. If the user deems the instructions inappropriate, additional support information or actions (e.g., "View more detailed guide" or "Contact support center") are suggested.

[1403] Gathering feedback

[1404] Users provide feedback on the information provided, for example, by rating it as "This information was very helpful." Because feedback helps improve the system, the device sends the feedback information to the server, which may include the results of the sentiment engine's analysis.

[1405] Saving feedback and retraining the AI

[1406] The server stores the received feedback information in a database. This feedback information is used to retrain the generative AI model and the emotion engine. This improves the accuracy of responses to subsequent search requests and the accuracy of emotion recognition. This system allows crew members to quickly and appropriately obtain necessary manual information, improving the quality of customer service and enabling flexible responses that take into account the user's emotional state.

[1407] For example, if the keyword "register operation new registration" is entered, the server will generate a manual including the operation procedure and return information such as, "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button." This entire process takes into account the user's emotional state and reflects it as appropriate feedback, thereby improving the overall performance of the system.

[1408] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1409] Step 1:

[1410] The server reads the connection settings and initializes the AI ​​model, emotion engine, and database connection. This process uses a Python-based Flask server, with TensorFlow or PyTorch used for the generated AI model, and OpenFace or EmotionAPI for the emotion engine.

[1411] Input: Connection configuration file

[1412] Output: System in initialization complete state

[1413] Specific actions: Read the server configuration file, initialize various libraries, and establish a database connection.

[1414] Step 2:

[1415] The user accesses the system and enters their username and password on the login screen. The terminal sends this authentication information to the server.

[1416] Input: Username, Password

[1417] Output: Session ID (if authentication is successful) or error message (if authentication fails)

[1418] Specific operation: The user enters information into the login form, and the device sends it to the server as a POST request. The server compares it with the database and generates a session ID.

[1419] Step 3:

[1420] The user enters keywords on the terminal and sends a search request. For example, they might enter "cash register operation new registration". The terminal converts these keywords into JSON format and sends it to the server.

[1421] Input: Keyword

[1422] Output: Search request (JSON format)

[1423] Specific operation: The user enters text into the search bar and presses the search button. The device uses the JavaScript JSON.stringify function to convert the keywords into JSON format and sends an AJAX request to the server.

[1424] Step 4:

[1425] The emotion engine analyzes the user's facial expressions and voice during input to recognize their emotional state. The analysis results are transmitted to the server. For example, if a user inputs with an unhappy facial expression or voice, the emotion engine will determine that they are "unhappy."

[1426] Input: Facial expression data, audio data

[1427] Output: Emotional state (e.g., "discomfort," "joy," etc.)

[1428] Specific operation: The device collects camera and microphone data in real time, performs sentiment analysis using OpenCV and TensorFlow models, and sends the results to the server in JSON format.

[1429] Step 5:

[1430] The server searches the database based on the received search request and sentiment analysis results. If no relevant documents are found, it uses a generative AI to generate new manual content.

[1431] Input: Search request, emotional state

[1432] Output: Formatted manual information

[1433] Specific operation: Executes SQL queries to search for relevant documents, and if necessary, sends prompts to generative AI models such as GPT-3 to generate new manual content. Adjusts tone and content based on emotional state.

[1434] Step 6:

[1435] The server sends formatted manual information to the terminal, which then displays it to the user.

[1436] Input: Formatted manual information

[1437] Output: Manual information displayed to the user

[1438] Specific operation: The server sends manual information to the terminal in JSON format, the terminal uses JavaScript to embed the data into HTML, and displays it to the user.

[1439] Step 7:

[1440] The user provides feedback on the information provided. For example, they might rate it as "This information was very helpful." The device sends this feedback, along with the sentiment engine's analysis results, to the server.

[1441] Input: Feedback, sentiment analysis results

[1442] Output: Sent feedback data

[1443] Specific operation: The user enters information into the feedback form and presses the submit button. The device sends the feedback data to the server in JSON format.

[1444] Step 8:

[1445] The server stores the received feedback information in a database. This feedback information is used to retrain generative AI models and emotion engines.

[1446] Input: Feedback data

[1447] Output: Saved feedback information, retrained AI model

[1448] Specific actions: Use SQL queries to save feedback data to the database, and then retrain the generative AI model and emotion engine using the new dataset.

[1449] (Application Example 2)

[1450] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1451] Conventional operational support systems in physical stores often fail to respond quickly and appropriately to problems faced by crew members, and a particular challenge is the lack of support that takes into account the emotional state of the crew. Furthermore, there are no systems that allow crew members to use advanced technologies such as voice input or smart glasses to support their work, which leads to decreased operational efficiency and negatively impacts the quality of customer service. This invention aims to improve operational efficiency and customer service quality by recognizing the emotional state of crew members and providing appropriate information.

[1452] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for reading connection settings and performing initialization, means for receiving authentication information from the user and performing authentication, means for receiving keywords from the user and processing search requests, means for searching for relevant documents from a database, means for generating new information using a generative AI if no relevant documents are found in the database, means for formatting the generated information and sending it to the user terminal, means for receiving feedback from the user and storing it in a database, means for retraining the generative AI using the received feedback, means for recognizing the user's emotional state using smart glasses, means for adjusting the tone of the generated information according to the user's emotional state, means for receiving questions from the user via voice input, and means for processing voice feedback. This enables the provision of appropriate information according to the emotional state of the crew and quick and effective work support using voice input.

[1453] "Connection settings" refer to the settings that include network configurations and authentication information necessary for each component of the system to communicate.

[1454] "Initialization" refers to the process of preparing the necessary settings and resources before a system starts operating, and ensuring that each component functions correctly.

[1455] User authentication is the process of verifying the authentication information (such as username and password) that a user enters to access a system, and confirming that they are a legitimate user.

[1456] A "keyword" refers to the words or phrases that users enter into the system as part of a search request.

[1457] A "search request" is a request made by a user to a system to retrieve specific information.

[1458] A "database" is a collection of data used by a system to manage and store information, enabling efficient information retrieval and updating.

[1459] "Related documents" refer to documents or guidelines that are directly related to what the user is looking for in their search request.

[1460] "Generative AI" is an artificial intelligence technology that automatically generates new information and documents based on user requests and emotional states.

[1461] "Formatting" refers to the process of adjusting the format and presentation of generated or retrieved information to make it easier for users to understand.

[1462] A "user terminal" refers to a device (such as a smartphone, tablet, or smart glasses) that a crew member uses to operate the system.

[1463] "Feedback" refers to the evaluations and comments that users give to the information provided or the function of the system.

[1464] "Retraining" is a learning process that uses collected feedback data to improve the performance of generative AI.

[1465] "Smart glasses" are wearable devices that users can wear and use to display and operate information visually and audibly.

[1466] "Emotional state" refers to the psychological state (e.g., joy, anger, fatigue, etc.) analyzed from the user's facial expressions and voice.

[1467] "Tone" refers to the tone of the language and expression used in the generated information, and it is adjusted according to the user's emotional state.

[1468] This invention is a system designed to support operations in retail stores, primarily using smart glasses to improve crew efficiency and customer service. The system consists of a server, user terminals (smart glasses), a generative AI model, an emotion engine, and a database.

[1469] Initialization and connection settings

[1470] The server loads connection settings and initializes the AI ​​model, emotion engine, and database connection when the system starts up. This process prepares the system so that each component can function correctly.

[1471] User Authentication

[1472] The user (crew member) wears smart glasses and authenticates by entering their username and password. The terminal sends this authentication information to the server, which then compares it with the database. If authentication is successful, the server generates a session ID and sends it back to the terminal.

[1473] Question input and sentiment recognition

[1474] When a user inputs a question by voice, the device sends the voice data to the server. Simultaneously, the smart glasses' camera and microphone are used to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's current emotional state.

[1475] Manual generation and information provision

[1476] The server uses a generative AI model to generate new information based on the received question and emotional state. It searches the database for relevant existing documents, and if no relevant documents are found, the generative AI model generates a new manual tailored to the user. The tone of the generated manual is adjusted to reflect the emotional state. The formatted information is displayed to the user through smart glasses.

[1477] Feedback gathering and retraining

[1478] The user provides voice feedback on the information provided. The device sends the feedback information to the server, which stores it in a database. The stored feedback information is used to retrain the generative AI model and emotion engine, improving the accuracy of future search requests and manual generation.

[1479] Hardware / software to use

[1480] Hardware:

[1481] Smart glasses (e.g., wearable devices)

[1482] server

[1483] software:

[1484] Flask (Web framework)

[1485] SQLite (database)

[1486] Emotion recognition engine

[1487] Generative AI models

[1488] Specific example

[1489] User authentication:

[1490] User: "Enter your username. Enter your password."

[1491] Question Input:

[1492] User: "Please tell me how to register a new product at the checkout."

[1493] Emotion recognition:

[1494] The emotion engine reads the user's tired facial expression and detects "fatigue."

[1495] Manual generation:

[1496] "To register a new product at the register, scan the product's barcode, enter the price and quantity, and then click the register button. If you need more detailed instructions, please press the support button."

[1497] feedback:

[1498] User: "This information was very helpful."

[1499] Example of a prompt

[1500] "The user asked 'How do I register a new product at the register?' and the emotion 'fatigue' was detected. Please provide detailed instructions in an appropriate tone."

[1501] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1502] Step 1:

[1503] The server reads the connection settings and performs the necessary initialization. This prepares the entire system for normal operation. Specifically, this includes initializing the AI ​​model and emotion engine, and establishing the database connection. The input is the configuration file, and the output is the system's initialized state.

[1504] Step 2:

[1505] The user enters their username and password using the smart glasses interface. The device sends this information to the server. The server compares the entered authentication information with its database, and if authentication is successful, generates a session ID and sends it back to the device. The input is the user authentication information, and the output is the session ID.

[1506] Step 3:

[1507] The user inputs a question by voice. The device sends the voice data to the server and simultaneously uses the smart glasses' camera and microphone to recognize the user's emotional state from their facial expressions and voice. The emotion engine analyzes this data to determine the user's emotional state. The input is voice data and facial expression data, and the output is the emotional state. The specific operation includes analysis by the emotion engine.

[1508] Step 4:

[1509] The server generates new information using a generative AI model based on the received question content and sentiment state. First, it searches the database for any relevant existing documents. If no relevant documents are found, the generative AI model generates a new manual using a prompt. The input is the question content and sentiment state, and the output is a formatted manual. The specific operation includes document generation by the generative AI model.

[1510] Step 5:

[1511] The generated manual has its tone adjusted according to the user's emotional state. The server sends this formatted information to the user's terminal (smart glasses). The terminal displays it to the user. The input is the unformatted manual information, and the output is the formatted manual displayed to the user. Specific actions include adjusting the tone of the information according to the user's emotions.

[1512] Step 6:

[1513] The user provides voice feedback on the information provided. The device sends the feedback information to the server. The server stores this information in a database. The input is voice feedback, and the output is the feedback information stored in the database.

[1514] Step 7:

[1515] The server uses stored feedback information to retrain the generative AI model and sentiment engine. This process improves the accuracy of subsequent search requests and manual generation. The input is feedback data, and the output is the retrained model. Specific actions include retraining the AI ​​model.

[1516] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1517] The data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One 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">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1518] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1519] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1520] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1521] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1522] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1523] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1524] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1525] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1526] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1527] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1528] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1529] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1530] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1531] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1532] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1533] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1534] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1535] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1536] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[1537] The following is further disclosed regarding the embodiments described above.

[1538] (Claim 1)

[1539] A means of loading and initializing connection settings,

[1540] A means of receiving authentication information from a user and performing authentication,

[1541] A means of receiving keywords from users and processing search requests,

[1542] A means of searching for relevant documents in a database,

[1543] If no documents related to the aforementioned database are found, a means for generating new information using a generative AI,

[1544] A means for formatting the generated information and sending it to the user terminal,

[1545] A means of receiving user feedback and saving it to a database,

[1546] A method for retraining generative AI using received feedback,

[1547] A system that includes this.

[1548] (Claim 2)

[1549] The system according to claim 1, wherein information generated based on a search request is automatically formatted and transmitted to the user terminal.

[1550] (Claim 3)

[1551] The system according to claim 1, further comprising means for retraining a generative AI to improve its performance using user-inputted feedback.

[1552] "Example 1"

[1553] (Claim 1)

[1554] A means of loading and initializing connection settings,

[1555] A means of receiving authentication information from a user and performing authentication,

[1556] A means of receiving keywords from users and processing search requests,

[1557] A means of searching for relevant documents in a database,

[1558] If no documents related to the aforementioned database are found, a means for generating new information using a generative AI model,

[1559] A means for formatting the generated information and sending it to the user terminal,

[1560] A means of receiving user feedback and saving it to a database,

[1561] A means of retraining a generative AI model using the received feedback,

[1562] A means for initializing the AI ​​model and database connection when the system starts up,

[1563] A means by which a user enters keywords into a search field and the device sends those keywords to the server,

[1564] A means of formatting the generated manual information and displaying it to the user,

[1565] A system that includes this.

[1566] (Claim 2)

[1567] The system according to claim 1, wherein information generated based on a search request is automatically formatted and transmitted to the user terminal.

[1568] (Claim 3)

[1569] The system according to claim 1, comprising means for retraining a generative AI model to improve its performance using user-inputted feedback.

[1570] "Application Example 1"

[1571] (Claim 1)

[1572] A means of loading and initializing connection settings,

[1573] A means of receiving authentication information from a user and performing authentication,

[1574] A means of receiving keywords from users and processing search requests,

[1575] A means of searching for relevant documents in a database,

[1576] If no documents related to the aforementioned database are found, a means for generating new information using a generative AI,

[1577] A means for formatting the generated information and sending it to the user terminal,

[1578] A means of receiving user feedback and saving it to a database,

[1579] A method for retraining generative AI using received feedback,

[1580] A means of providing manuals regarding driving operations and system settings on information terminals and personal mobile devices of autonomous vehicles,

[1581] A system that includes this.

[1582] (Claim 2)

[1583] The system according to claim 1, wherein information generated based on a search request is automatically formatted and transmitted to the user terminal.

[1584] (Claim 3)

[1585] The system according to claim 1, further comprising means for retraining a generative AI to improve its performance using user-inputted feedback.

[1586] "Example 2 of combining an emotion engine"

[1587] (Claim 1)

[1588] A means of loading and initializing connection settings,

[1589] A means of receiving authentication information from a user and performing authentication,

[1590] A means of receiving keywords from users and processing search requests,

[1591] A means of analyzing the user's facial expressions and voice during input to recognize their emotional state,

[1592] A means of searching for relevant documents in a database,

[1593] If no documents related to the aforementioned database are found, a means for generating new information using a generative AI,

[1594] A means for formatting the generated information based on the user's emotional state and sending it to the user's terminal,

[1595] A means of receiving user feedback and saving it to a database,

[1596] A method for retraining generative AI using received feedback,

[1597] A system that includes this.

[1598] (Claim 2)

[1599] The system according to claim 1, wherein information generated based on a search request is automatically formatted and transmitted to the user terminal.

[1600] (Claim 3)

[1601] The system according to claim 1, further comprising means for performing sentiment analysis when receiving feedback from a user, saving the results as feedback, and retraining the generative AI to improve its performance.

[1602] "Application example 2 when combining with an emotional engine"

[1603] (Claim 1)

[1604] A means of loading and initializing connection settings,

[1605] A means of receiving authentication information from a user and performing authentication,

[1606] A means of receiving keywords from users and processing search requests,

[1607] A means of searching for relevant documents in a database,

[1608] If no documents related to the aforementioned database are found, a means for generating new information using a generative AI,

[1609] A means for formatting the generated information and sending it to the user terminal,

[1610] A means of receiving user feedback and saving it to a database,

[1611] A method for retraining generative AI using received feedback,

[1612] A means of recognizing a user's emotional state using smart glasses,

[1613] A means of adjusting the tone of information generated according to the user's emotional state,

[1614] A means of receiving questions from users via voice input,

[1615] A means of processing audio feedback,

[1616] A system that includes this.

[1617] (Claim 2)

[1618] The system according to claim 1, wherein information generated based on a search request is automatically formatted and transmitted to the user terminal.

[1619] (Claim 3)

[1620] The system according to claim 1, further comprising means for retraining a generative AI to improve its performance using user-inputted feedback. [Explanation of symbols]

[1621] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of loading and initializing connection settings, A means of receiving authentication information from a user and performing authentication, A means of receiving keywords from users and processing search requests, A means of searching for relevant documents in a database, If no documents related to the aforementioned database are found, a means for generating new information using a generative AI, A means for formatting the generated information and sending it to the user terminal, A means of receiving user feedback and saving it to a database, A method for retraining generative AI using received feedback, A system that includes this.

2. The system according to claim 1, wherein information generated based on a search request is automatically formatted and transmitted to the user terminal.

3. The system according to claim 1, further comprising means for retraining a generative AI to improve its performance using user-inputted feedback.

Images