System

Abstract

A system is provided.SOLUTION: A system comprising: means for a user to select a particular character; means for loading a AI model corresponding to the selected character; means for receiving and analyzing incoming messages from the user; means for the AI model to generate an appropriate response based on the analysis; means for providing the generated response to the user; means for collecting feedback provided by the user; and means for tuning the AI model based on the collected feedback.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Many users today want to communicate with specific characters from anime or games, but existing technology makes it difficult to faithfully reproduce the character's personality and worldview. This leaves users dissatisfied with the interaction experience. Furthermore, there is no established method for properly collecting feedback to improve the performance of AI models, resulting in a lack of accuracy and naturalness in responses. [Means for solving the problem]

[0005] The present invention provides a means for a user to select a specific character and load an AI model corresponding to that character. It then includes a means for analyzing the input message from the user and having the AI ​​model generate an appropriate response based on the results. Furthermore, by providing the generated response to the user and providing a means for collecting evaluations of the interaction experience, the AI ​​model can be tuned and its performance improved based on the feedback. Furthermore, by including a means for storing user evaluations in a evaluation database and analyzing the data, continuous and effective model improvement is possible.

[0006] "User" refers to an individual or organization that uses the target system.

[0007] "Character" refers to a person or creature that appears in fictional works such as anime and games.

[0008] "AI model" refers to a collection of mathematical or computational algorithms designed to solve a specific problem using artificial intelligence techniques.

[0009] "Loading" refers to reading specific data or programs into memory in a computer system.

[0010] "Message" refers to information such as text or voice that a user inputs to the system.

[0011] "Analysis" refers to the computational process of understanding input data and grasping its meaning and intent.

[0012] "Response" refers to the reply or reaction that the system gives to a message from the user.

[0013] "Feedback" refers to the evaluation or opinion a user provides about an interactive experience.

[0014] "Tuning" refers to adjustments and improvements made to improve a system's performance and responsiveness.

[0015] A "database" refers to a system that systematically stores and manages large amounts of data.

[0016] "Review Database" means a database used to store and later analyze user reviews and feedback.

[0017] "Analysis results" refers to the information or output obtained after analyzing input data. [Brief explanation of the drawings]

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

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

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

[0021] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0022] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

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

[0024] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0026] [First embodiment]

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

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

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

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

[0031] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

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

[0033] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

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

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

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

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

[0039] This invention provides a platform for users to interact with specific characters, and includes the following processes: the user selects a character, loads an AI model corresponding to the selected character, analyzes messages from the user, the AI ​​model generates an appropriate response, and provides that response to the user. Furthermore, the system has the function of collecting feedback from users and tuning the AI ​​model based on that feedback.

[0040] Application launch and authentication

[0041] 1. User: Launch the "Minkuri PF" application.

[0042] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[0043] 3. User: After entering the authentication information, click the login button.

[0044] 4. Device: Sends authentication information to the server.

[0045] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[0046] Character Selection

[0047] 1. Device: Display a list of selectable characters on the home screen.

[0048] 2. User: Select the character you want to interact with.

[0049] 3. Device: Send the selected character ID to the server.

[0050] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[0051] Starting a conversation

[0052] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[0053] 2. User: Enter the message you want to communicate and press the send button.

[0054] 3. Terminal: Sends a message to the server.

[0055] 4. Server: Analyzes the received message and inputs it into the AI ​​model.

[0056] 5. Server: Sends the appropriate response generated by the AI ​​model to the device.

[0057] 6. Terminal: Displays the sent response on an interactive screen.

[0058] Specific examples

[0059] User: Type "Hello, Character A!" and submit.

[0060] Terminal: Sends a message to the server.

[0061] Server: Analyzes the message content and has the AI ​​model generate a response to the "hello" part.

[0062] Server: Generates a response saying "Hello, I'm Character A!" and sends it to the device.

[0063] Terminal: Display "Hello, I'm Character A!" on the dialogue screen.

[0064] Gathering feedback

[0065] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0066] 2. User: Select and submit a rating.

[0067] 3. Device: Sends the selected rating to the server.

[0068] 4. Server: Collects feedback and stores it in a reputation database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[0069] Tuning the model

[0070] 1. Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[0071] 2. Server: Identify areas for improvement and retrain the AI ​​model to improve performance if necessary.

[0072] 3. Server: Apply the improved model to the platform and provide it to users.

[0073] In this way, the system of the present invention allows for natural interaction with characters and continuously improves the performance of the model.

[0074] The processing flow will be explained below.

[0075] Step 1:

[0076] User: Launch the "Minkuri PF" application.

[0077] Terminal: Displays a login screen to the user and prompts them to enter their authentication information (username, password).

[0078] Step 2:

[0079] User: Enters authentication information and clicks the login button.

[0080] Terminal: Sends the entered authentication information to the server.

[0081] Step 3:

[0082] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[0083] Device: Receives a successful authentication message and displays the home screen.

[0084] Step 4:

[0085] Device: The home screen displays a list of characters that the user can choose from.

[0086] User: Select the character you want to interact with.

[0087] Step 5:

[0088] Device: Sends the selected character ID to the server.

[0089] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[0090] Step 6:

[0091] Terminal: Displays an interactive screen and prompts the user to enter a message.

[0092] User: Enter a message and press send.

[0093] Step 7:

[0094] Terminal: Sends the entered message to the server.

[0095] Server: Parses the message and extracts important keywords and context.

[0096] Step 8:

[0097] Server: Based on the analysis results, the AI ​​model generates an appropriate response.

[0098] Server: Generates and sends the response to the device.

[0099] Step 9:

[0100] Terminal: Displays the received response on the interactive screen.

[0101] User: Type your next message and continue the conversation.

[0102] Step 10:

[0103] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0104] User: Select and submit a rating.

[0105] Step 11:

[0106] Device: Sends the user-selected rating to the server.

[0107] Server: Receives ratings and stores them in a feedback database.

[0108] Step 12:

[0109] Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[0110] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance.

[0111] Step 13:

[0112] Server: Applying the improved AI model to the platform and providing it to users.

[0113] Example 1

[0114] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0115] Conventional dialogue systems struggle to provide consistent responses when users interact with specific characters. Furthermore, feedback collection and AI model tuning are often done manually, resulting in slow performance improvements. Furthermore, insufficient user authentication processes create security challenges.

[0116] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0117] In this invention, the server includes means for inputting user authentication information and starting a session if authentication is successful, means for the user to select a specific character, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for the AI ​​model to generate an appropriate response based on the analysis result, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback, thereby making it possible to provide consistent and advanced responses, rapidly improve the model based on feedback, and achieve strong user authentication security.

[0118] "User authentication information" is information used to identify a user and verify access rights.

[0119] A "session" is a unit for recording and tracking a series of operations performed while a user is accessing a system.

[0120] A "character" is a virtual agent that interacts with the user within a dialogue system.

[0121] An "AI model" is an algorithm or system that uses artificial intelligence techniques to generate appropriate responses to given inputs.

[0122] An "input message" is information in text, voice, or other form that a user sends to a dialogue system.

[0123] "Parsing" is the process of understanding a received input message and identifying its meaning or intent.

[0124] A "response" is a reply message generated by an AI model in response to an input message.

[0125] "Feedback" refers to the ratings and opinions that users provide about their experience using the system.

[0126] "Tuning" is the process of improving the performance of an AI model based on collected feedback.

[0127] This invention provides a platform for users to interact with specific characters. The system allows users to select a character, loads an AI model corresponding to the selected character, analyzes messages from the user, and the AI ​​model generates an appropriate response and provides that response to the user. Furthermore, the system has the function of collecting user feedback and tuning the AI ​​model based on that feedback.

[0128] The hardware used includes servers (e.g., Dell PowerEdge R740) and user-operated devices (e.g., smartphones, tablets).The software used includes application software (e.g., "Minkuri PF") and AI model software (e.g., GPT-3, BERT).

[0129] Application launch and authentication

[0130] The user launches the Minkuri PF application on their device. The launched application displays a login screen and prompts the user to enter authentication information (user name and password).

[0131] When the user enters authentication information and presses the login button, the device sends the authentication information to the server. The server verifies whether the authentication information is correct, and if authentication is successful, starts a session and instructs the device to display the home screen.

[0132] Character Selection

[0133] The device will display a list of selectable characters on the home screen.

[0134] The user selects the character they want to interact with, and the device sends the selected character's ID to the server, which loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[0135] Starting a conversation

[0136] The terminal displays an interactive screen and prompts the user to input a message.

[0137] The user inputs a message they wish to communicate and presses the send button, and the terminal sends the message to the server.

[0138] The server analyzes the received message and inputs it into the AI ​​model, which generates an appropriate response that the server then sends to the device.

[0139] The terminal displays the transmitted response on an interactive screen.

[0140] As a concrete example, consider a scenario where a user types and sends "Hello, Character A!" In this case, the server analyzes the message content and has the AI ​​model generate an appropriate response for the "Hello" part. The generated response is "Hello, I'm Character A!", which is sent to the device and displayed on the interactive screen.

[0141] Gathering feedback

[0142] After the interaction is completed, the terminal displays a screen asking the user to rate the interaction experience (e.g., five stars).

[0143] The user selects and submits a rating. The device then sends the selected rating to the server. The server collects the feedback and stores it in a rating database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[0144] Tuning the model

[0145] The server periodically analyzes the feedback data to evaluate the performance of the AI ​​model, finds areas for improvement, and retrains the AI ​​model to improve its performance as needed. The improved model is then applied to the platform and provided to users.

[0146] In this way, our system allows for natural interaction with characters and can continuously improve the model's performance based on feedback.

[0147] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0148] Step 1: Launch the Application and Authenticate

[0149] The user launches the application on the device and is presented with a login screen where they can enter their username and password.

[0150] The user enters authentication information (user name, password) and presses the login button.

[0151] The terminal sends the entered authentication information to the server. At this time, the authentication information is sent in encrypted form.

[0152] The server checks the authentication information against the database, and if it matches, it sends a response indicating successful authentication to the terminal. If it does not match, it sends a message indicating unsuccessful authentication.

[0153] Input: Username and Password

[0154] Output: Session ID if authentication is successful, error message if authentication fails

[0155] What it does: Send and verify credentials

[0156] Step 2: Character Selection

[0157] After successful authentication, the device will display a list of selectable characters on the home screen.

[0158] The user selects the character with which they want to interact.

[0159] The terminal transmits the selected character ID to the server.

[0160] Input: User's character selection

[0161] Output: Selected character ID

[0162] Action: Display and select a character from the list

[0163] Step 3: Loading the AI ​​model

[0164] The server loads the AI ​​model corresponding to the received character ID.

[0165] The server notifies the device that the AI ​​model has finished loading.

[0166] Input: Character ID

[0167] Output: Notification that the AI ​​model is ready to load

[0168] Action: Loading an AI model

[0169] Step 4: Start a conversation

[0170] The terminal displays an interactive screen and prompts the user to input a message.

[0171] The user inputs a message to be communicated and presses the send button.

[0172] The terminal transmits the input message to the server.

[0173] Input: User's message

[0174] Output: Message sent to server

[0175] Action: Type and send a message

[0176] Step 5: Parsing the message and generating a response

[0177] The server analyzes the received message and inputs its contents into the AI ​​model.

[0178] The server receives the response generated by the AI ​​model and sends it to the device.

[0179] Input: User's message

[0180] Output: Response from the AI ​​model

[0181] Action: Parse the message and generate a response

[0182] Step 6: Display the response

[0183] The terminal displays the response received from the server on an interactive screen.

[0184] Input: Response from the server

[0185] Output: Response display on the interactive screen

[0186] Action: Display response

[0187] Step 7: Gather feedback

[0188] After the interaction is completed, the terminal displays a screen requesting the user to evaluate the interaction experience.

[0189] The user selects and submits a rating.

[0190] The terminal transmits the selected rating to the server.

[0191] Input: User rating

[0192] Output: Send rating to server

[0193] Action: Enter and submit a rating

[0194] Step 8: Storing and analyzing feedback

[0195] The server stores the received feedback in a ratings database.

[0196] The server analyzes the feedback data and uses it to improve the performance of the AI ​​model.

[0197] Input: User feedback

[0198] Output: Save to database and analysis results

[0199] How it works: Saving and analyzing feedback

[0200] Step 9: Tune the AI ​​model

[0201] The server periodically analyzes the feedback data and retrains the AI ​​model as needed to improve its performance.

[0202] The server applies the improved model to the platform and provides it to the user.

[0203] Input: Feedback data

[0204] Output: Improved model

[0205] How it works: Retraining and applying AI models

[0206] The system's processing allows for natural interaction with characters and allows for continuous performance improvement.

[0207] (Application example 1)

[0208] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0209] In today's brick-and-mortar stores, many customers want an environment where they can easily ask questions about detailed product information and recommended products. However, with limited staff numbers, it is difficult to respond to all customer inquiries quickly and appropriately. There is also a lack of efficient ways to collect customer feedback and reflect it in service improvements. As a result, customer satisfaction may decline, which could have a negative impact on store sales. An effective way to solve this problem is needed.

[0210] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0211] In this invention, the server includes a means for a user to select a specific character, a means for loading an AI model corresponding to the selected character, a means for receiving and analyzing an input message from the user, and a means for applying the generated response to product introductions and inquiries in a physical store. This allows for automated dialogue with customers, enabling prompt and appropriate responses. Furthermore, tuning the AI ​​model based on collected feedback enables continuous improvement of the service.

[0212] "User" refers to any individual or corporation that uses this system.

[0213] "Character" refers to a virtual person or character that a user selects as a target for interaction.

[0214] "AI Model" refers to the artificial intelligence algorithms and data models that generate responses to selected characters.

[0215] "Input message" refers to text or voice information sent by a user to the system.

[0216] "Parsing" refers to understanding the content of an input message and processing it to generate an appropriate response.

[0217] "Response" refers to the reply or information provided to the user generated by the AI ​​model.

[0218] "Feedback" refers to the ratings and opinions provided by users about their interactive experiences.

[0219] "Tuning" refers to making adjustments to improve the performance and accuracy of an AI model based on collected feedback.

[0220] "Brick and mortar store" refers to a physical location for selling products.

[0221] "Product introduction" refers to explaining the features and benefits of a product to customers in a physical store.

[0222] "Inquiry response" refers to responding to customer questions and requests and providing appropriate information and services.

[0223] The present invention includes a system that provides a platform for users to interact with specific characters and effectively introduces products and responds to inquiries in physical stores.

[0224] The server first receives user authentication information and allows the user to log in to the system. After authentication, the user is asked to select from multiple characters and the corresponding AI model is loaded based on the selection. The server then receives the input message from the user, analyzes the message, and generates an appropriate response. The generated response is used in the physical store to introduce products and respond to inquiries. In addition, feedback provided by the user after the interaction is collected and used to tune the AI ​​model, thereby continuously improving the quality of service.

[0225] The main hardware used is the device used by the user, such as a smartphone or smart glasses, while the server side requires computing resources to run a high-performance AI model. This is achieved by using cloud computing services (e.g., Amazon Web Services or Google Cloud Platform). On the software side, Flask is used to provide an API, and data communication is often in JSON format.

[0226] Specifically, the user launches the app on their smartphone, logs in, and then selects the "Shopping Assist Character." An example of the prompt is shown below.

[0227] User: "What are the features of this product?"

[0228] AI: "This product incorporates the latest technology and combines functionality with design."

[0229] Through these interactions, customers can quickly and accurately obtain the information they are looking for. After the interaction is over, the user provides a star rating, which is sent as feedback to the server. The server then uses this feedback to retrain the AI ​​model and improve the overall system so that it can provide more appropriate responses in the next interaction.

[0230] By introducing this system, customer service in physical stores will be automated, and it is expected that customer satisfaction will improve and store operations will become more efficient.

[0231] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0232] Step 1: User selects a specific character

[0233] Input: The user launches the app and enters their login information (username, password).

[0234] Data processing: The device sends the entered login information to the server, which checks the authentication information against its database.

[0235] Output: If authentication is successful, a list of characters the user can choose from is displayed on the terminal.

[0236] Step 2: Load the AI ​​model corresponding to the selected character

[0237] Input: The user selects a character. For example, "Shopping Assist Character."

[0238] Data processing: The device sends the selected character ID to the server, which then loads the AI ​​model corresponding to the character ID into memory.

[0239] Output: Notifies that a character has been selected and displays the dialogue screen on the terminal.

[0240] Step 3: Receive and parse the user input message

[0241] Input: The user types a message into the interactive screen and sends it. For example, "What are the features of this product?"

[0242] Data processing: The device sends the input message to the server, which analyzes the message using techniques such as morphological analysis to extract meaning.

[0243] Output: Based on the analysis results, a prompt is generated for the AI ​​model, ready to generate an appropriate response.

[0244] Step 4: The AI ​​model generates an appropriate response

[0245] Input: Parsed message content (prompt sentence). For example, "User: What are the features of this product?"

[0246] Data processing: The server inputs the analysis results into the AI ​​model, and the generative AI model generates a response.

[0247] Output: The generated response, for example, "This product uses the latest technology and combines functionality and design."

[0248] Step 5: Providing the generated response to the user

[0249] Input: The response sentence generated by the AI ​​model.

[0250] Data processing: The server sends the generated response to the terminal, which then displays the received response on the interactive screen.

[0251] Output: A response to be displayed on the interactive screen. For example, "This product uses the latest technology and combines functionality and design."

[0252] Step 6: Gather user feedback

[0253] Input: A feedback request screen after the interaction is completed. The user can enter and submit an evaluation of their interaction experience.

[0254] Data processing: The device sends the entered rating to the server, which stores the rating in the rating database.

[0255] Output: The evaluation data is saved in a database.

[0256] Step 7: Tune the AI ​​model based on collected feedback

[0257] Input: Collected feedback data.

[0258] Data processing: The server analyzes the feedback data and evaluates the performance of the AI ​​model, retraining it as needed and tuning the model.

[0259] Output: A new, improved AI model that will enable a better response the next time you interact with it.

[0260] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0261] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[0262] Application launch and authentication

[0263] 1. User: Launch the "Minkuri PF" application.

[0264] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[0265] 3. User: After entering the authentication information, click the login button.

[0266] 4. Device: Sends authentication information to the server.

[0267] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[0268] Character Selection

[0269] 1. Device: Display a list of characters that the user can select from on the home screen.

[0270] 2. User: Select the character you want to interact with.

[0271] 3. Device: Send the selected character ID to the server.

[0272] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[0273] Dialogue initiation and emotion recognition

[0274] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[0275] 2. User: Enter a message and press the send button.

[0276] 3. Terminal: Sends the entered message to the server.

[0277] 4. Server: Receives the message and detects the user's emotion using the emotion engine.

[0278] 5. Server: Add emotional information to the message content analysis results and input it into the AI ​​model.

[0279] 6. Server: Based on the emotional information, the AI ​​model generates an appropriate response and sends it to the device.

[0280] 7. Terminal: Displays the received response on the interactive screen.

[0281] Specific examples

[0282] User: Type "I'm really tired today" and submit.

[0283] Terminal: Sends a message to the server.

[0284] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[0285] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[0286] Server: Generates and sends the response to the device.

[0287] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[0288] Gathering feedback

[0289] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0290] 2. User: Select and submit a rating.

[0291] 3. Device: Sends user ratings to the server.

[0292] 4. Server: Stores the ratings in a feedback database and then analyzes the data to improve the performance of the AI ​​model.

[0293] Tuning the model

[0294] 1. Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[0295] 2. Server: Identify areas for improvement and retrain the AI ​​model as needed to improve performance.

[0296] 3. Server: Apply the improved AI model to the platform and provide it to users.

[0297] In this way, the system of the present invention, combined with emotion recognition capabilities, can achieve more natural and personalized interactions, providing responses based on the user's emotions, improving the interaction experience and increasing user satisfaction.

[0298] The processing flow will be explained below.

[0299] Step 1:

[0300] User: Launch the "Minkuri PF" application.

[0301] Terminal: Displays a login screen and prompts the user to enter authentication information (username, password).

[0302] Step 2:

[0303] User: Enters authentication information and clicks the login button.

[0304] Terminal: Sends the entered authentication information to the server.

[0305] Step 3:

[0306] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[0307] Device: Receives a successful authentication message and displays the home screen.

[0308] Step 4:

[0309] Device: The home screen displays a list of characters that the user can choose from.

[0310] User: Select the character you want to interact with.

[0311] Step 5:

[0312] Device: Sends the selected character ID to the server.

[0313] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[0314] Step 6:

[0315] Terminal: Displays an interactive screen and prompts the user to enter a message.

[0316] User: Enter a message and press send.

[0317] Step 7:

[0318] Terminal: Sends the entered message to the server.

[0319] Server: Receives the message and detects the user's emotion using the emotion engine.

[0320] Step 8:

[0321] Server: The analysis results include the emotional information detected by the emotion engine along with the message content.

[0322] Step 9:

[0323] Server: The AI ​​model generates an appropriate response taking into account emotional information.

[0324] Server: Generates and sends the response to the device.

[0325] Step 10:

[0326] Terminal: Display the response on the interactive screen.

[0327] User: Continue the conversation by entering a next message if desired.

[0328] Specific examples of emotion recognition

[0329] User: Type "I'm really tired today" and submit.

[0330] Terminal: Sends a message to the server.

[0331] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[0332] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[0333] Server: Generates and sends the response to the device.

[0334] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[0335] Step 11:

[0336] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0337] User: Select and submit a rating.

[0338] Step 12:

[0339] Device: Sends user ratings to the server.

[0340] Server: Stores the evaluations in a feedback database, then analyzes the data and uses it to improve the performance of the AI ​​model.

[0341] Step 13:

[0342] Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[0343] Server: Retrains the AI ​​model as needed to improve its performance.

[0344] Server: Applying the improved AI model to the platform and providing it to users.

[0345] Example 2

[0346] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0347] Conventional dialogue systems often provide simple responses without understanding the user's emotions. This results in low-quality dialogue and makes it difficult to improve user satisfaction. Furthermore, they lack the means to effectively utilize feedback to improve the system, making it difficult to respond quickly to user needs.

[0348] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0349] In this invention, the server includes: a means for a user to select a specific character; a means for loading an AI model corresponding to the selected character; a means for receiving and analyzing an input message from the user; a means for the AI ​​model to generate an appropriate response based on the analysis result; a means for providing the generated response to the user; a means for collecting feedback provided by the user; a means for adjusting the AI ​​model based on the collected feedback; a means for analyzing the input message using an emotion engine to recognize the user's emotions; and a means for the AI ​​model to generate a response based on the emotion information. This enables the provision of higher-quality dialogue based on the user's emotions, thereby improving user satisfaction. Furthermore, the AI ​​model can be continuously adjusted and improved using feedback, enabling the system to quickly respond to user needs.

[0350] "User" refers to a human being who uses a dialogue system.

[0351] A "character" is a visual or conceptual representation of a person or object that is the subject of a dialogue, and is used to interact with the user in a dialogue system.

[0352] An "artificial intelligence model" refers to an algorithm that learns from large amounts of data to generate and analyze text, and specifically includes generative AI models (e.g., GPT-3).

[0353] An "input message" refers to information such as text or voice that a user sends to a dialogue system.

[0354] "Analysis" refers to the act of using natural language processing technology to understand the meaning and emotions of an input message and process the information.

[0355] "Response" refers to a reply or message to the user that the artificial intelligence model generates based on the analysis results.

[0356] An "emotion engine" refers to an algorithm or system for detecting and classifying a user's emotions from an input message.

[0357] "Feedback" refers to information such as evaluations and opinions that users provide to a dialogue system.

[0358] "Review Database" refers to a database for storing and managing feedback collected from users.

[0359] "Tuning" refers to the act of retraining or changing parameters to improve the performance or behavior of an AI model based on collected feedback and other data.

[0360] MODE FOR CARRYING OUT THE INVENTION

[0361] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[0362] First, the user launches an application on a device such as a smartphone, tablet, or PC. This application provides the interface necessary for the interactive system. Specific application names include "Minkuri PF."

[0363] When a user launches an application, the device displays a login screen and prompts for a username and password. The user enters the authentication information and presses the login button, and the device sends the authentication information to the server. The server verifies whether the received authentication information is correct, and if so, starts a session and instructs the device to display the home screen.

[0364] When the home screen is displayed, the device presents the user with a list of characters to choose from. The user selects the character they want to interact with, and the device sends the selected character's ID to the server. The server loads the AI ​​model (e.g., GPT-3) corresponding to the character ID and notifies the device that it is ready.

[0365] When a conversation begins, the device displays a dialogue screen and prompts the user to enter a message. The user enters a message and presses the send button, and the device sends the message to the server. The server analyzes the received message using an emotion engine to recognize the user's emotions. By analyzing specific keywords and sentences, emotions can be classified as "joy," "sadness," "anger," "fatigue," etc.

[0366] Based on the analysis results, the server provides emotion information to the generative AI model, which then generates an appropriate response. The response is sent from the server to the device, which then displays it on the interactive screen. For example, if a user types, "I'm very tired today," the emotion engine detects the keyword "tired" and recognizes the emotion as "fatigue." The generative AI model then generates a response such as, "Good work! Take a good rest," which is displayed on the device.

[0367] After the interaction is completed, the device displays a screen asking the user to rate the interaction experience, and the user enters and submits the rating. The device then sends the rating to the server, which stores it in a feedback database. The feedback is periodically analyzed and used to improve the AI ​​model.

[0368] The following are examples of prompt sentences:

[0369] The user types, "I'm very tired today." Recognize the user's emotions and generate an appropriate response.

[0370] To implement this system, the following hardware and software are required.

[0371] Hardware: devices such as smartphones, tablets, and computers

[0372] Software: conversational applications (e.g., Minkuri PF), server software, generative AI models (e.g., GPT-3), emotion recognition engines (e.g., NLTK, SpaCy)

[0373] In this way, the system of the present invention combines emotion recognition capabilities to enable more natural and personalized interactions, improving user satisfaction.

[0374] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0375] Step 1: Launch the Application and Authenticate

[0376] User: Launch the "Minkuri PF" application on a device such as a smartphone, tablet, or PC.

[0377] Terminal: Displays the login screen and prompts for a username and password. Receives the authentication information (username and password) entered by the user.

[0378] User: Enter your username and password and click the login button.

[0379] Terminal: Sends the entered information to the server. The input is (username, password) and the output is (sending authentication information).

[0380] Server: Verifies whether the received authentication information matches the data in the database. Verification involves performing a data check to compare the authentication information with the database. The output is (authentication result).

[0381] Server: Once authenticated, it starts a session and instructs the device to display the home screen. The output is (instruction to display home screen).

[0382] Step 2: Character Selection

[0383] Device: Displays the home screen and provides the user with a list of characters to choose from.

[0384] User: Select the character you want to interact with and tap on that character.

[0385] Terminal: Sends the selected character's ID to the server. Input is (selected character ID), output is (sent character ID).

[0386] Server: Based on the received character ID, load the corresponding AI model (e.g., GPT-3). The input is (character ID) and the output is (loaded AI model).

[0387] Server: Notify the terminal that loading is complete. The output is (notification of readiness).

[0388] Step 3: Initiating a dialogue and recognizing emotions

[0389] Terminal: Displays an interactive screen and prompts the user to enter a message.

[0390] User: Enter a message and press send.

[0391] Terminal: Sends the entered message to the server. The input is (user's message) and the output is (sent message).

[0392] Server: Analyzes received messages and detects user emotions using an emotion engine. The emotion engine uses natural language processing technology (e.g., NLTK, SpaCy) to analyze keywords in messages and assign emotion labels. The input is (message) and the output is (emotional information).

[0393] Server: The analysis results, including emotional information, are input into the generative AI model, which then generates an appropriate response. The input is (the analysis results, including emotional information), and the output is (the generated response).

[0394] Server: Sends the generated response to the terminal. The output is (sent response).

[0395] Terminal: Displays the received response on an interactive screen. The input is (the generated response) and the output is (the display of the response).

[0396] Step 4: Gather feedback

[0397] Terminal: After the interaction is completed, an evaluation screen is displayed, asking the user to rate their interaction experience.

[0398] User: Enter a rating and click the submit button.

[0399] Terminal: Sends the entered rating to the server. The input is (user rating) and the output is (sent rating).

[0400] Server: Stores the received ratings in a feedback database. The input is (user ratings) and the output is (storage of feedback data).

[0401] Step 5: Tune the model

[0402] Server: Periodically analyzes feedback data and emotion data to evaluate the performance of the AI ​​model. Performance evaluation uses statistical analysis and machine learning algorithms based on the collected data. The input is (feedback data, emotion data) and the output is (performance evaluation results).

[0403] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance. The input is (performance evaluation results) and the output is (an improved AI model).

[0404] Server: Apply the improved AI model to the platform and provide it to the user. The output is (application of the improved model).

[0405] (Application example 2)

[0406] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0407] Conventional self-driving vehicles lack the ability to recognize the emotional state of the driver and passengers in real time and provide appropriate responses or suggestions based on that information. This poses a problem in that it is not possible to reduce the mental and physical burden on the driver and passengers during long driving periods or stressful driving environments. Furthermore, there is a lack of a mechanism for utilizing feedback based on the user's emotions to improve the performance of AI models. This invention aims to solve these problems and improve the user experience in self-driving vehicles.

[0408] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0409] In this invention, the server includes a means for capturing a user's voice in real time and converting the voice to text, a means for analyzing the user's emotions using an emotion engine based on the analysis results, and a means for an AI model to generate an appropriate response based on the analysis results and emotion information. This enables appropriate responses and suggestions to be made according to the user's emotional state. Furthermore, by collecting feedback provided by users and continuously tuning the AI ​​model based on that feedback, the system's performance can be improved.

[0410] A "character" refers to a virtual entity that a user interacts with and that is visually displayed.

[0411] "AI model" refers to an algorithm for data analysis and response generation built using artificial intelligence.

[0412] "Emotion engine" refers to software or algorithms for analyzing messages received from a user and recognizing the user's emotions.

[0413] "Audio capture" refers to the process of collecting a user's voice using a microphone or the like.

[0414] "Text-to-text" refers to the process of converting audio data obtained through voice capture into written information.

[0415] "Response generation" refers to the process by which an AI model creates an appropriate response to a user based on analysis results and emotional information.

[0416] "Feedback" refers to evaluations and opinions provided by users regarding their experience using the system.

[0417] "Tuning" refers to the process of improving and adjusting the performance of an AI model based on collected feedback.

[0418] "Server" refers to a computer system that processes and analyzes various types of data.

[0419] "Real-time" refers to data processing and response generation occurring almost simultaneously with the passage of real time.

[0420] The present invention is a system for improving the user experience in an autonomous vehicle, and specific embodiments thereof are described below.

[0421] System Configuration

[0422] The system includes means for selecting a specific character with which a user will interact, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for analyzing the user's emotions using an emotion engine based on the analysis result, means for the AI ​​model to generate an appropriate response based on the analysis result and the emotion information, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback.

[0423] Hardware and Software

[0424] Hardware

[0425] Smart glasses or smartphone: Used for user voice input and feedback collection.

[0426] Server: Performs data processing and analysis, and hosts AI models.

[0427] software

[0428] Speech recognition engine (Google Speech Recognition API): Converts speech collected from a microphone into text.

[0429] Emotion analysis model (Hugging Face's Transformers library): Analyzes transcribed speech data to identify emotions.

[0430] Text-to-speech engine (pyttsx3): Provides responses generated by the AI ​​model as audio feedback to the user.

[0431] Data processing flow

[0432] First, the user uses smart glasses or a smartphone to input voice. This voice is converted into text through a speech recognition engine and sent to the server. The server then uses an emotion engine to analyze the emotion of the input text. Based on the analysis results and emotion information, the AI ​​model generates an appropriate response and provides feedback to the user as voice through a text-to-speech engine.

[0433] Specific examples

[0434] Input prompt example

[0435] "I'm very tired today."

[0436] "I'm tired from the long journey"

[0437] In response, the emotion engine identifies the emotion "fatigue," and the AI ​​model generates a response such as, "Thank you for your hard work. I'll play some relaxing music. Please take a short break at the next rest point." If the user provides feedback on this response, that feedback is collected and stored by the server and used to tune the AI ​​model in the future.

[0438] System Features

[0439] This system can provide appropriate responses and suggestions in real time according to the user's emotional state. For example, if the user feels tired while driving, it will automatically play relaxing music and guide them to appropriate rest stops. This helps reduce fatigue and stress caused by long driving hours.

[0440] The above is a specific embodiment of the present invention. Use of this system is expected to significantly improve the user experience in autonomous vehicles.

[0441] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0442] Step 1:

[0443] A user uses smart glasses or a smartphone to provide voice input.

[0444] Input: User's voice

[0445] Output: Audio data

[0446] Action: The user speaks to the device, saying something like "I'm very tired today."

[0447] Step 2:

[0448] The device uses a speech recognition engine (Google Speech Recognition API) to convert the voice data into text.

[0449] Input: Audio data

[0450] Output: Text data

[0451] How it works: The device's microphone captures audio and sends it to the Google Speech Recognition API, which converts it into text: "I'm very tired today."

[0452] Step 3:

[0453] The terminal transmits the text data to the server.

[0454] Input: Text data

[0455] Output: Request sent to server

[0456] Action: The device sends the converted text "I'm very tired today" to the server.

[0457] Step 4:

[0458] The server analyzes the emotions in the text data using an emotion engine (Hugging Face's Transformers library).

[0459] Input: Text data

[0460] Output: Emotional information (e.g., "fatigue")

[0461] How it works: The server inputs text data into the emotion engine and identifies the emotion "fatigue" from the keyword "tired."

[0462] Step 5:

[0463] The server uses an AI model to generate an appropriate response based on the analysis results and emotional information.

[0464] Input: Emotion information and text data

[0465] Output: Response text (e.g. "Good work! We'll play some relaxing music. Please take a short break at the next break point.")

[0466] How it works: Based on the emotional information "fatigue," the server uses an AI model to generate an appropriate response, which is then output as text.

[0467] Step 6:

[0468] The server sends the generated response text to the terminal.

[0469] Input: Response text

[0470] Output: Sending a request to the terminal

[0471] Behavior: The server generates a response "Good work! We'll play some relaxing music. Please take a short break at the next rest point." and sends it to the device.

[0472] Step 7:

[0473] The device uses a text-to-speech engine (pyttsx3) to audibly output the response text.

[0474] Input: Response text

[0475] Output: Audio data

[0476] How it works: The device inputs the response text into a text-to-speech engine and provides it as audio feedback to the user.

[0477] Step 8:

[0478] The user provides feedback on the response.

[0479] Input: User ratings and opinions

[0480] Output: Feedback data

[0481] How it works: The user enters a rating or opinion on the device.

[0482] Step 9:

[0483] The terminal transmits the feedback data to the server.

[0484] Input: Feedback data

[0485] Output: Request sent to server

[0486] Operation: The terminal transmits the entered evaluation data to the server.

[0487] Step 10:

[0488] The server stores the collected feedback data and uses it to tune the AI ​​model.

[0489] Input: Feedback data

[0490] Output: An improved AI model

[0491] How it works: The server stores the feedback data in a database and periodically analyzes the data to improve the performance of the AI ​​model.

[0492] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0493] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0494] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0495] [Second embodiment]

[0496] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0497] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0498] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0499] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0500] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0501] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0502] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0503] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0504] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0505] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0506] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0507] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0508] This invention provides a platform for users to interact with specific characters, and includes the following processes: the user selects a character, loads an AI model corresponding to the selected character, analyzes messages from the user, the AI ​​model generates an appropriate response, and provides that response to the user. Furthermore, the system has the function of collecting feedback from users and tuning the AI ​​model based on that feedback.

[0509] Application launch and authentication

[0510] 1. User: Launch the "Minkuri PF" application.

[0511] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[0512] 3. User: After entering the authentication information, click the login button.

[0513] 4. Device: Sends authentication information to the server.

[0514] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[0515] Character Selection

[0516] 1. Device: Display a list of selectable characters on the home screen.

[0517] 2. User: Select the character you want to interact with.

[0518] 3. Device: Send the selected character ID to the server.

[0519] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[0520] Starting a conversation

[0521] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[0522] 2. User: Enter the message you want to communicate and press the send button.

[0523] 3. Terminal: Sends a message to the server.

[0524] 4. Server: Analyzes the received message and inputs it into the AI ​​model.

[0525] 5. Server: Sends the appropriate response generated by the AI ​​model to the device.

[0526] 6. Terminal: Displays the sent response on an interactive screen.

[0527] Specific examples

[0528] User: Type "Hello, Character A!" and submit.

[0529] Terminal: Sends a message to the server.

[0530] Server: Analyzes the message content and has the AI ​​model generate a response to the "hello" part.

[0531] Server: Generates a response saying "Hello, I'm Character A!" and sends it to the device.

[0532] Terminal: Display "Hello, I'm Character A!" on the dialogue screen.

[0533] Gathering feedback

[0534] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0535] 2. User: Select and submit a rating.

[0536] 3. Device: Sends the selected rating to the server.

[0537] 4. Server: Collects feedback and stores it in a reputation database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[0538] Tuning the model

[0539] 1. Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[0540] 2. Server: Identify areas for improvement and retrain the AI ​​model to improve performance if necessary.

[0541] 3. Server: Apply the improved model to the platform and provide it to users.

[0542] In this way, the system of the present invention allows for natural interaction with characters and continuously improves the performance of the model.

[0543] The processing flow will be explained below.

[0544] Step 1:

[0545] User: Launch the "Minkuri PF" application.

[0546] Terminal: Displays a login screen to the user and prompts them to enter their authentication information (username, password).

[0547] Step 2:

[0548] User: Enters authentication information and clicks the login button.

[0549] Terminal: Sends the entered authentication information to the server.

[0550] Step 3:

[0551] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[0552] Device: Receives a successful authentication message and displays the home screen.

[0553] Step 4:

[0554] Device: The home screen displays a list of characters that the user can choose from.

[0555] User: Select the character you want to interact with.

[0556] Step 5:

[0557] Device: Sends the selected character ID to the server.

[0558] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[0559] Step 6:

[0560] Terminal: Displays an interactive screen and prompts the user to enter a message.

[0561] User: Enter a message and press send.

[0562] Step 7:

[0563] Terminal: Sends the entered message to the server.

[0564] Server: Parses the message and extracts important keywords and context.

[0565] Step 8:

[0566] Server: Based on the analysis results, the AI ​​model generates an appropriate response.

[0567] Server: Generates and sends the response to the device.

[0568] Step 9:

[0569] Terminal: Displays the received response on the interactive screen.

[0570] User: Type your next message and continue the conversation.

[0571] Step 10:

[0572] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0573] User: Select and submit a rating.

[0574] Step 11:

[0575] Device: Sends the user-selected rating to the server.

[0576] Server: Receives ratings and stores them in a feedback database.

[0577] Step 12:

[0578] Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[0579] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance.

[0580] Step 13:

[0581] Server: Applying the improved AI model to the platform and providing it to users.

[0582] Example 1

[0583] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0584] Conventional dialogue systems struggle to provide consistent responses when users interact with specific characters. Furthermore, feedback collection and AI model tuning are often done manually, resulting in slow performance improvements. Furthermore, insufficient user authentication processes create security challenges.

[0585] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0586] In this invention, the server includes means for inputting user authentication information and starting a session if authentication is successful, means for the user to select a specific character, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for the AI ​​model to generate an appropriate response based on the analysis result, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback, thereby making it possible to provide consistent and advanced responses, rapidly improve the model based on feedback, and achieve strong user authentication security.

[0587] "User authentication information" is information used to identify a user and verify access rights.

[0588] A "session" is a unit for recording and tracking a series of operations performed while a user is accessing a system.

[0589] A "character" is a virtual agent that interacts with the user within a dialogue system.

[0590] An "AI model" is an algorithm or system that uses artificial intelligence techniques to generate appropriate responses to given inputs.

[0591] An "input message" is information in text, voice, or other form that a user sends to a dialogue system.

[0592] "Parsing" is the process of understanding a received input message and identifying its meaning or intent.

[0593] A "response" is a reply message generated by an AI model in response to an input message.

[0594] "Feedback" refers to the ratings and opinions that users provide about their experience using the system.

[0595] "Tuning" is the process of improving the performance of an AI model based on collected feedback.

[0596] This invention provides a platform for users to interact with specific characters. The system allows users to select a character, loads an AI model corresponding to the selected character, analyzes messages from the user, and the AI ​​model generates an appropriate response and provides that response to the user. Furthermore, the system has the function of collecting user feedback and tuning the AI ​​model based on that feedback.

[0597] The hardware used includes servers (e.g., Dell PowerEdge R740) and user-operated devices (e.g., smartphones, tablets).The software used includes application software (e.g., "Minkuri PF") and AI model software (e.g., GPT-3, BERT).

[0598] Application launch and authentication

[0599] The user launches the Minkuri PF application on their device. The launched application displays a login screen and prompts the user to enter authentication information (user name and password).

[0600] When the user enters authentication information and presses the login button, the device sends the authentication information to the server. The server verifies whether the authentication information is correct, and if authentication is successful, starts a session and instructs the device to display the home screen.

[0601] Character Selection

[0602] The device will display a list of selectable characters on the home screen.

[0603] The user selects the character they want to interact with, and the device sends the selected character's ID to the server, which loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[0604] Starting a conversation

[0605] The terminal displays an interactive screen and prompts the user to input a message.

[0606] The user inputs a message they wish to communicate and presses the send button, and the terminal sends the message to the server.

[0607] The server analyzes the received message and inputs it into the AI ​​model, which generates an appropriate response that the server then sends to the device.

[0608] The terminal displays the transmitted response on an interactive screen.

[0609] As a concrete example, consider a scenario where a user types and sends "Hello, Character A!" In this case, the server analyzes the message content and has the AI ​​model generate an appropriate response for the "Hello" part. The generated response is "Hello, I'm Character A!", which is sent to the device and displayed on the interactive screen.

[0610] Gathering feedback

[0611] After the interaction is completed, the terminal displays a screen asking the user to rate the interaction experience (e.g., five stars).

[0612] The user selects and submits a rating. The device then sends the selected rating to the server. The server collects the feedback and stores it in a rating database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[0613] Tuning the model

[0614] The server periodically analyzes the feedback data to evaluate the performance of the AI ​​model, finds areas for improvement, and retrains the AI ​​model to improve its performance as needed. The improved model is then applied to the platform and provided to users.

[0615] In this way, our system allows for natural interaction with characters and can continuously improve the model's performance based on feedback.

[0616] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0617] Step 1: Launch the Application and Authenticate

[0618] The user launches the application on the device and is presented with a login screen where they can enter their username and password.

[0619] The user enters authentication information (user name, password) and presses the login button.

[0620] The terminal sends the entered authentication information to the server. At this time, the authentication information is sent in encrypted form.

[0621] The server checks the authentication information against the database, and if it matches, it sends a response indicating successful authentication to the terminal. If it does not match, it sends a message indicating unsuccessful authentication.

[0622] Input: Username and Password

[0623] Output: Session ID if authentication is successful, error message if authentication fails

[0624] What it does: Send and verify credentials

[0625] Step 2: Character Selection

[0626] After successful authentication, the device will display a list of selectable characters on the home screen.

[0627] The user selects the character with which they want to interact.

[0628] The terminal transmits the selected character ID to the server.

[0629] Input: User's character selection

[0630] Output: Selected character ID

[0631] Action: Display and select a character from the list

[0632] Step 3: Loading the AI ​​model

[0633] The server loads the AI ​​model corresponding to the received character ID.

[0634] The server notifies the device that the AI ​​model has finished loading.

[0635] Input: Character ID

[0636] Output: Notification that the AI ​​model is ready to load

[0637] Action: Loading an AI model

[0638] Step 4: Start a conversation

[0639] The terminal displays an interactive screen and prompts the user to input a message.

[0640] The user inputs a message to be communicated and presses the send button.

[0641] The terminal transmits the input message to the server.

[0642] Input: User's message

[0643] Output: Message sent to server

[0644] Action: Enter and send a message

[0645] Step 5: Parsing the message and generating a response

[0646] The server analyzes the received message and inputs its contents into the AI ​​model.

[0647] The server receives the response generated by the AI ​​model and sends it to the device.

[0648] Input: User's message

[0649] Output: Response from the AI ​​model

[0650] Action: Parse the message and generate a response

[0651] Step 6: Display the response

[0652] The terminal displays the response received from the server on an interactive screen.

[0653] Input: Response from the server

[0654] Output: Response display on the interactive screen

[0655] Action: Display response

[0656] Step 7: Gather feedback

[0657] After the interaction is completed, the terminal displays a screen requesting the user to evaluate the interaction experience.

[0658] The user selects and submits a rating.

[0659] The terminal transmits the selected rating to the server.

[0660] Input: User rating

[0661] Output: Send rating to server

[0662] Action: Enter and submit a rating

[0663] Step 8: Storing and analyzing feedback

[0664] The server stores the received feedback in a ratings database.

[0665] The server analyzes the feedback data and uses it to improve the performance of the AI ​​model.

[0666] Input: User feedback

[0667] Output: Save to database and analysis results

[0668] How it works: Saving and analyzing feedback

[0669] Step 9: Tune the AI ​​model

[0670] The server periodically analyzes the feedback data and retrains the AI ​​model as needed to improve its performance.

[0671] The server applies the improved model to the platform and provides it to the user.

[0672] Input: Feedback data

[0673] Output: Improved model

[0674] How it works: Retraining and applying AI models

[0675] The system's processing allows for natural interaction with characters and allows for continuous performance improvement.

[0676] (Application example 1)

[0677] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0678] In today's brick-and-mortar stores, many customers want an environment where they can easily ask questions about detailed product information and recommended products. However, with limited staff numbers, it is difficult to respond to all customer inquiries quickly and appropriately. There is also a lack of efficient ways to collect customer feedback and reflect it in service improvements. As a result, customer satisfaction may decline, which could have a negative impact on store sales. An effective way to solve this problem is needed.

[0679] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0680] In this invention, the server includes a means for a user to select a specific character, a means for loading an AI model corresponding to the selected character, a means for receiving and analyzing an input message from the user, and a means for applying the generated response to product introductions and inquiries in a physical store. This allows for automated dialogue with customers, enabling prompt and appropriate responses. Furthermore, tuning the AI ​​model based on collected feedback enables continuous improvement of the service.

[0681] "User" refers to any individual or corporation that uses this system.

[0682] "Character" refers to a virtual person or character that a user selects as a target for interaction.

[0683] "AI Model" refers to the artificial intelligence algorithms and data models that generate responses to selected characters.

[0684] "Input message" refers to text or voice information sent by a user to the system.

[0685] "Parsing" refers to understanding the content of an input message and processing it to generate an appropriate response.

[0686] "Response" refers to the reply or information provided to the user generated by the AI ​​model.

[0687] "Feedback" refers to the ratings and opinions provided by users about their interactive experiences.

[0688] "Tuning" refers to making adjustments to improve the performance and accuracy of an AI model based on collected feedback.

[0689] "Brick and mortar store" refers to a physical location for selling products.

[0690] "Product introduction" refers to explaining the features and benefits of a product to customers in a physical store.

[0691] "Inquiry response" refers to responding to customer questions and requests and providing appropriate information and services.

[0692] The present invention includes a system that provides a platform for users to interact with specific characters and effectively introduces products and responds to inquiries in physical stores.

[0693] The server first receives user authentication information and allows the user to log in to the system. After authentication, the user is asked to select from multiple characters and the corresponding AI model is loaded based on the selection. The server then receives the input message from the user, analyzes the message, and generates an appropriate response. The generated response is used in the physical store to introduce products and respond to inquiries. In addition, feedback provided by the user after the interaction is collected and used to tune the AI ​​model, thereby continuously improving the quality of service.

[0694] The main hardware used is the device used by the user, such as a smartphone or smart glasses, while the server side requires computing resources to run a high-performance AI model. This is achieved by using cloud computing services (e.g., Amazon Web Services or Google Cloud Platform). On the software side, Flask is used to provide an API, and data communication is often in JSON format.

[0695] Specifically, the user launches the app on their smartphone, logs in, and then selects the "Shopping Assist Character." An example of the prompt is shown below.

[0696] User: "What are the features of this product?"

[0697] AI: "This product incorporates the latest technology and combines functionality with design."

[0698] Through these interactions, customers can quickly and accurately obtain the information they are looking for. After the interaction is over, the user provides a star rating, which is sent as feedback to the server. The server then uses this feedback to retrain the AI ​​model and improve the overall system so that it can provide more appropriate responses in the next interaction.

[0699] By introducing this system, customer service in physical stores will be automated, and it is expected that customer satisfaction will improve and store operations will become more efficient.

[0700] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0701] Step 1: User selects a specific character

[0702] Input: The user launches the app and enters their login information (username, password).

[0703] Data processing: The device sends the entered login information to the server, which checks the authentication information against its database.

[0704] Output: If authentication is successful, a list of characters the user can choose from is displayed on the terminal.

[0705] Step 2: Load the AI ​​model corresponding to the selected character

[0706] Input: The user selects a character. For example, "Shopping Assist Character."

[0707] Data processing: The device sends the selected character ID to the server, which then loads the AI ​​model corresponding to the character ID into memory.

[0708] Output: Notifies that a character has been selected and displays the dialogue screen on the terminal.

[0709] Step 3: Receive and parse the user input message

[0710] Input: The user types a message into the interactive screen and sends it. For example, "What are the features of this product?"

[0711] Data processing: The device sends the input message to the server, which analyzes the message using techniques such as morphological analysis to extract meaning.

[0712] Output: Based on the analysis results, a prompt is generated for the AI ​​model, ready to generate an appropriate response.

[0713] Step 4: The AI ​​model generates an appropriate response

[0714] Input: Parsed message content (prompt sentence). For example, "User: What are the features of this product?"

[0715] Data processing: The server inputs the analysis results into the AI ​​model, and the generative AI model generates a response.

[0716] Output: The generated response, for example, "This product uses the latest technology and combines functionality and design."

[0717] Step 5: Providing the generated response to the user

[0718] Input: The response sentence generated by the AI ​​model.

[0719] Data processing: The server sends the generated response to the terminal, which then displays the received response on the interactive screen.

[0720] Output: A response to be displayed on the interactive screen. For example, "This product uses the latest technology and combines functionality and design."

[0721] Step 6: Gather user feedback

[0722] Input: A feedback request screen after the interaction is completed. The user can enter and submit an evaluation of their interaction experience.

[0723] Data processing: The device sends the entered rating to the server, which stores the rating in the rating database.

[0724] Output: The evaluation data is saved in a database.

[0725] Step 7: Tune the AI ​​model based on collected feedback

[0726] Input: Collected feedback data.

[0727] Data processing: The server analyzes the feedback data and evaluates the performance of the AI ​​model, retraining it as needed and tuning the model.

[0728] Output: A new, improved AI model that will enable a better response the next time you interact with it.

[0729] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0730] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[0731] Application launch and authentication

[0732] 1. User: Launch the "Minkuri PF" application.

[0733] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[0734] 3. User: After entering the authentication information, click the login button.

[0735] 4. Device: Sends authentication information to the server.

[0736] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[0737] Character Selection

[0738] 1. Device: Display a list of characters that the user can select from on the home screen.

[0739] 2. User: Select the character you want to interact with.

[0740] 3. Device: Send the selected character ID to the server.

[0741] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[0742] Dialogue initiation and emotion recognition

[0743] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[0744] 2. User: Enter a message and press the send button.

[0745] 3. Terminal: Sends the entered message to the server.

[0746] 4. Server: Receives the message and detects the user's emotion using the emotion engine.

[0747] 5. Server: Add emotional information to the message content analysis results and input it into the AI ​​model.

[0748] 6. Server: Based on the emotional information, the AI ​​model generates an appropriate response and sends it to the device.

[0749] 7. Terminal: Displays the received response on the interactive screen.

[0750] Specific examples

[0751] User: Type "I'm really tired today" and submit.

[0752] Terminal: Sends a message to the server.

[0753] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[0754] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[0755] Server: Generates and sends the response to the device.

[0756] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[0757] Gathering feedback

[0758] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0759] 2. User: Select and submit a rating.

[0760] 3. Device: Sends user ratings to the server.

[0761] 4. Server: Stores the ratings in a feedback database and then analyzes the data to improve the performance of the AI ​​model.

[0762] Tuning the model

[0763] 1. Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[0764] 2. Server: Identify areas for improvement and retrain the AI ​​model as needed to improve performance.

[0765] 3. Server: Apply the improved AI model to the platform and provide it to users.

[0766] In this way, the system of the present invention, combined with emotion recognition capabilities, can achieve more natural and personalized interactions, providing responses based on the user's emotions, improving the interaction experience and increasing user satisfaction.

[0767] The processing flow will be explained below.

[0768] Step 1:

[0769] User: Launch the "Minkuri PF" application.

[0770] Terminal: Displays a login screen and prompts the user to enter authentication information (username, password).

[0771] Step 2:

[0772] User: Enters authentication information and clicks the login button.

[0773] Terminal: Sends the entered authentication information to the server.

[0774] Step 3:

[0775] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[0776] Device: Receives a successful authentication message and displays the home screen.

[0777] Step 4:

[0778] Device: The home screen displays a list of characters that the user can choose from.

[0779] User: Select the character you want to interact with.

[0780] Step 5:

[0781] Device: Sends the selected character ID to the server.

[0782] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[0783] Step 6:

[0784] Terminal: Displays an interactive screen and prompts the user to enter a message.

[0785] User: Enter a message and press send.

[0786] Step 7:

[0787] Terminal: Sends the entered message to the server.

[0788] Server: Receives the message and detects the user's emotion using the emotion engine.

[0789] Step 8:

[0790] Server: The analysis results include the emotional information detected by the emotion engine along with the message content.

[0791] Step 9:

[0792] Server: The AI ​​model generates an appropriate response taking into account emotional information.

[0793] Server: Generates and sends the response to the device.

[0794] Step 10:

[0795] Terminal: Display the response on the interactive screen.

[0796] User: Continue the conversation by entering a next message if desired.

[0797] Specific examples of emotion recognition

[0798] User: Type "I'm really tired today" and submit.

[0799] Terminal: Sends a message to the server.

[0800] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[0801] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[0802] Server: Generates and sends the response to the device.

[0803] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[0804] Step 11:

[0805] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[0806] User: Select and submit a rating.

[0807] Step 12:

[0808] Device: Sends user ratings to the server.

[0809] Server: Stores the evaluations in a feedback database, then analyzes the data and uses it to improve the performance of the AI ​​model.

[0810] Step 13:

[0811] Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[0812] Server: Retrains the AI ​​model as needed to improve its performance.

[0813] Server: Applying the improved AI model to the platform and providing it to users.

[0814] Example 2

[0815] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0816] Conventional dialogue systems often provide simple responses without understanding the user's emotions. This results in low-quality dialogue and makes it difficult to improve user satisfaction. Furthermore, they lack the means to effectively utilize feedback to improve the system, making it difficult to respond quickly to user needs.

[0817] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0818] In this invention, the server includes: a means for a user to select a specific character; a means for loading an AI model corresponding to the selected character; a means for receiving and analyzing an input message from the user; a means for the AI ​​model to generate an appropriate response based on the analysis result; a means for providing the generated response to the user; a means for collecting feedback provided by the user; a means for adjusting the AI ​​model based on the collected feedback; a means for analyzing the input message using an emotion engine to recognize the user's emotions; and a means for the AI ​​model to generate a response based on the emotion information. This enables the provision of higher-quality dialogue based on the user's emotions, thereby improving user satisfaction. Furthermore, the AI ​​model can be continuously adjusted and improved using feedback, enabling the system to quickly respond to user needs.

[0819] "User" refers to a human being who uses a dialogue system.

[0820] A "character" is a visual or conceptual representation of a person or object that is the subject of a dialogue, and is used to interact with the user in a dialogue system.

[0821] An "artificial intelligence model" refers to an algorithm that learns from large amounts of data to generate and analyze text, and specifically includes generative AI models (e.g., GPT-3).

[0822] An "input message" refers to information such as text or voice that a user sends to a dialogue system.

[0823] "Analysis" refers to the act of using natural language processing technology to understand the meaning and emotions of an input message and process the information.

[0824] "Response" refers to a reply or message to the user that the artificial intelligence model generates based on the analysis results.

[0825] An "emotion engine" refers to an algorithm or system for detecting and classifying a user's emotions from an input message.

[0826] "Feedback" refers to information such as evaluations and opinions that users provide to a dialogue system.

[0827] "Review Database" refers to a database for storing and managing feedback collected from users.

[0828] "Tuning" refers to the act of retraining or changing parameters to improve the performance or behavior of an AI model based on collected feedback and other data.

[0829] MODE FOR CARRYING OUT THE INVENTION

[0830] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[0831] First, the user launches an application on a device such as a smartphone, tablet, or PC. This application provides the interface necessary for the interactive system. Specific application names include "Minkuri PF."

[0832] When a user launches an application, the device displays a login screen and prompts for a username and password. The user enters the authentication information and presses the login button, and the device sends the authentication information to the server. The server verifies whether the received authentication information is correct, and if so, starts a session and instructs the device to display the home screen.

[0833] When the home screen is displayed, the device presents the user with a list of characters to choose from. The user selects the character they want to interact with, and the device sends the selected character's ID to the server. The server loads the AI ​​model (e.g., GPT-3) corresponding to the character ID and notifies the device that it is ready.

[0834] When a conversation begins, the device displays a dialogue screen and prompts the user to enter a message. The user enters a message and presses the send button, and the device sends the message to the server. The server analyzes the received message using an emotion engine to recognize the user's emotions. By analyzing specific keywords and sentences, emotions can be classified as "joy," "sadness," "anger," "fatigue," etc.

[0835] Based on the analysis results, the server provides emotion information to the generative AI model, which then generates an appropriate response. The response is sent from the server to the device, which then displays it on the interactive screen. For example, if a user types, "I'm very tired today," the emotion engine detects the keyword "tired" and recognizes the emotion as "fatigue." The generative AI model then generates a response such as, "Good work! Take a good rest," which is displayed on the device.

[0836] After the interaction is completed, the device displays a screen asking the user to rate the interaction experience, and the user enters and submits the rating. The device then sends the rating to the server, which stores it in a feedback database. The feedback is periodically analyzed and used to improve the AI ​​model.

[0837] The following are examples of prompt sentences:

[0838] The user types, "I'm very tired today." Recognize the user's emotions and generate an appropriate response.

[0839] To implement this system, the following hardware and software are required.

[0840] Hardware: devices such as smartphones, tablets, and computers

[0841] Software: conversational applications (e.g., Minkuri PF), server software, generative AI models (e.g., GPT-3), emotion recognition engines (e.g., NLTK, SpaCy)

[0842] In this way, the system of the present invention combines emotion recognition capabilities to enable more natural and personalized interactions, improving user satisfaction.

[0843] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0844] Step 1: Launch the Application and Authenticate

[0845] User: Launch the "Minkuri PF" application on a device such as a smartphone, tablet, or PC.

[0846] Terminal: Displays the login screen and prompts for a username and password. Receives the authentication information (username and password) entered by the user.

[0847] User: Enter your username and password and click the login button.

[0848] Terminal: Sends the entered information to the server. The input is (username, password) and the output is (sending authentication information).

[0849] Server: Verifies whether the received authentication information matches the data in the database. Verification involves performing a data check to compare the authentication information with the database. The output is (authentication result).

[0850] Server: Once authenticated, it starts a session and instructs the device to display the home screen. The output is (instruction to display home screen).

[0851] Step 2: Character Selection

[0852] Device: Displays the home screen and provides the user with a list of characters to choose from.

[0853] User: Select the character you want to interact with and tap on that character.

[0854] Terminal: Sends the selected character's ID to the server. Input is (selected character ID), output is (sent character ID).

[0855] Server: Based on the received character ID, load the corresponding AI model (e.g., GPT-3). The input is (character ID) and the output is (loaded AI model).

[0856] Server: Notify the terminal that loading is complete. The output is (notification of readiness).

[0857] Step 3: Initiating a dialogue and recognizing emotions

[0858] Terminal: Displays an interactive screen and prompts the user to enter a message.

[0859] User: Enter a message and press send.

[0860] Terminal: Sends the entered message to the server. The input is (user's message) and the output is (sent message).

[0861] Server: Analyzes received messages and detects user emotions using an emotion engine. The emotion engine uses natural language processing technology (e.g., NLTK, SpaCy) to analyze keywords in messages and assign emotion labels. The input is (message) and the output is (emotional information).

[0862] Server: The analysis results, including emotional information, are input into the generative AI model, which then generates an appropriate response. The input is (the analysis results, including emotional information), and the output is (the generated response).

[0863] Server: Sends the generated response to the terminal. The output is (sent response).

[0864] Terminal: Displays the received response on an interactive screen. The input is (the generated response) and the output is (the display of the response).

[0865] Step 4: Gather feedback

[0866] Terminal: After the interaction is completed, an evaluation screen is displayed, asking the user to rate their interaction experience.

[0867] User: Enter a rating and click the submit button.

[0868] Terminal: Sends the entered rating to the server. The input is (user rating) and the output is (sent rating).

[0869] Server: Stores the received ratings in a feedback database. The input is (user ratings) and the output is (storage of feedback data).

[0870] Step 5: Tune the model

[0871] Server: Periodically analyzes feedback data and emotion data to evaluate the performance of the AI ​​model. Performance evaluation uses statistical analysis and machine learning algorithms based on the collected data. The input is (feedback data, emotion data) and the output is (performance evaluation results).

[0872] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance. The input is (performance evaluation results) and the output is (an improved AI model).

[0873] Server: Apply the improved AI model to the platform and provide it to the user. The output is (application of the improved model).

[0874] (Application example 2)

[0875] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0876] Conventional self-driving vehicles lack the ability to recognize the emotional state of the driver and passengers in real time and provide appropriate responses or suggestions based on that information. This poses a problem in that it is not possible to reduce the mental and physical burden on the driver and passengers during long driving periods or stressful driving environments. Furthermore, there is a lack of a mechanism for utilizing feedback based on the user's emotions to improve the performance of AI models. This invention aims to solve these problems and improve the user experience in self-driving vehicles.

[0877] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0878] In this invention, the server includes a means for capturing a user's voice in real time and converting the voice to text, a means for analyzing the user's emotions using an emotion engine based on the analysis results, and a means for an AI model to generate an appropriate response based on the analysis results and emotion information. This enables appropriate responses and suggestions to be made according to the user's emotional state. Furthermore, by collecting feedback provided by users and continuously tuning the AI ​​model based on that feedback, the system's performance can be improved.

[0879] A "character" refers to a virtual entity that a user interacts with and that is visually displayed.

[0880] "AI model" refers to an algorithm for data analysis and response generation built using artificial intelligence.

[0881] "Emotion engine" refers to software or algorithms for analyzing messages received from a user and recognizing the user's emotions.

[0882] "Audio capture" refers to the process of collecting a user's voice using a microphone or the like.

[0883] "Text-to-text" refers to the process of converting audio data obtained through voice capture into written information.

[0884] "Response generation" refers to the process by which an AI model creates an appropriate response to a user based on analysis results and emotional information.

[0885] "Feedback" refers to evaluations and opinions provided by users regarding their experience using the system.

[0886] "Tuning" refers to the process of improving and adjusting the performance of an AI model based on collected feedback.

[0887] "Server" refers to a computer system that processes and analyzes various types of data.

[0888] "Real-time" refers to data processing and response generation occurring almost simultaneously with the passage of real time.

[0889] The present invention is a system for improving the user experience in an autonomous vehicle, and specific embodiments thereof are described below.

[0890] System Configuration

[0891] The system includes means for selecting a specific character with which a user will interact, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for analyzing the user's emotions using an emotion engine based on the analysis result, means for the AI ​​model to generate an appropriate response based on the analysis result and the emotion information, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback.

[0892] Hardware and Software

[0893] Hardware

[0894] Smart glasses or smartphone: Used for user voice input and feedback collection.

[0895] Server: Performs data processing and analysis, and hosts AI models.

[0896] software

[0897] Speech recognition engine (Google Speech Recognition API): Converts speech collected from a microphone into text.

[0898] Emotion analysis model (Hugging Face's Transformers library): Analyzes transcribed speech data to identify emotions.

[0899] Text-to-speech engine (pyttsx3): Provides responses generated by the AI ​​model as audio feedback to the user.

[0900] Data processing flow

[0901] First, the user uses smart glasses or a smartphone to input voice. This voice is converted into text through a speech recognition engine and sent to the server. The server then uses an emotion engine to analyze the emotion of the input text. Based on the analysis results and emotion information, the AI ​​model generates an appropriate response and provides feedback to the user as voice through a text-to-speech engine.

[0902] Specific examples

[0903] Input prompt statement example

[0904] "I'm very tired today."

[0905] "I'm tired from the long journey"

[0906] In response, the emotion engine identifies the emotion "fatigue," and the AI ​​model generates a response such as, "Thank you for your hard work. I'll play some relaxing music. Please take a short break at the next rest point." If the user provides feedback on this response, that feedback is collected and stored by the server and used to tune the AI ​​model in the future.

[0907] System Features

[0908] This system can provide appropriate responses and suggestions in real time according to the user's emotional state. For example, if the user feels tired while driving, it will automatically play relaxing music and guide them to appropriate rest stops. This helps reduce fatigue and stress caused by long driving hours.

[0909] The above is a specific embodiment of the present invention. Use of this system is expected to significantly improve the user experience in autonomous vehicles.

[0910] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0911] Step 1:

[0912] A user uses smart glasses or a smartphone to provide voice input.

[0913] Input: User's voice

[0914] Output: Audio data

[0915] Action: The user speaks to the device, saying something like "I'm very tired today."

[0916] Step 2:

[0917] The device uses a speech recognition engine (Google Speech Recognition API) to convert the voice data into text.

[0918] Input: Audio data

[0919] Output: Text data

[0920] How it works: The device's microphone captures audio and sends it to the Google Speech Recognition API, which converts it into text: "I'm very tired today."

[0921] Step 3:

[0922] The terminal transmits the text data to the server.

[0923] Input: Text data

[0924] Output: Request sent to server

[0925] Action: The device sends the converted text "I'm very tired today" to the server.

[0926] Step 4:

[0927] The server analyzes the emotions in the text data using an emotion engine (Hugging Face's Transformers library).

[0928] Input: Text data

[0929] Output: Emotional information (e.g., "fatigue")

[0930] How it works: The server inputs text data into the emotion engine and identifies the emotion "fatigue" from the keyword "tired."

[0931] Step 5:

[0932] The server uses an AI model to generate an appropriate response based on the analysis results and emotional information.

[0933] Input: Emotion information and text data

[0934] Output: Response text (e.g. "Good work! We'll play some relaxing music. Please take a short break at the next break point.")

[0935] How it works: Based on the emotional information "fatigue," the server uses an AI model to generate an appropriate response, which is then output as text.

[0936] Step 6:

[0937] The server sends the generated response text to the terminal.

[0938] Input: Response text

[0939] Output: Sending a request to the terminal

[0940] Behavior: The server generates a response "Good work! We'll play some relaxing music. Please take a short break at the next rest point." and sends it to the device.

[0941] Step 7:

[0942] The device uses a text-to-speech engine (pyttsx3) to audibly output the response text.

[0943] Input: Response text

[0944] Output: Audio data

[0945] How it works: The device inputs the response text into a text-to-speech engine and provides it as audio feedback to the user.

[0946] Step 8:

[0947] The user provides feedback on the response.

[0948] Input: User ratings and opinions

[0949] Output: Feedback data

[0950] How it works: The user enters a rating or opinion on the device.

[0951] Step 9:

[0952] The terminal transmits the feedback data to the server.

[0953] Input: Feedback data

[0954] Output: Request sent to server

[0955] Operation: The terminal transmits the entered evaluation data to the server.

[0956] Step 10:

[0957] The server stores the collected feedback data and uses it to tune the AI ​​model.

[0958] Input: Feedback data

[0959] Output: An improved AI model

[0960] How it works: The server stores the feedback data in a database and periodically analyzes the data to improve the performance of the AI ​​model.

[0961] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0962] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0963] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0964] [Third embodiment]

[0965] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0966] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0967] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0968] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0969] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0970] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0971] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0972] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0973] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0974] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0975] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0976] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0977] This invention provides a platform for users to interact with specific characters, and includes the following processes: the user selects a character, loads an AI model corresponding to the selected character, analyzes messages from the user, the AI ​​model generates an appropriate response, and provides that response to the user. Furthermore, the system has the function of collecting feedback from users and tuning the AI ​​model based on that feedback.

[0978] Application launch and authentication

[0979] 1. User: Launch the "Minkuri PF" application.

[0980] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[0981] 3. User: After entering the authentication information, click the login button.

[0982] 4. Device: Sends authentication information to the server.

[0983] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[0984] Character Selection

[0985] 1. Device: Display a list of selectable characters on the home screen.

[0986] 2. User: Select the character you want to interact with.

[0987] 3. Device: Send the selected character ID to the server.

[0988] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[0989] Starting a conversation

[0990] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[0991] 2. User: Enter the message you want to communicate and press the send button.

[0992] 3. Terminal: Sends a message to the server.

[0993] 4. Server: Analyzes the received message and inputs it into the AI ​​model.

[0994] 5. Server: Sends the appropriate response generated by the AI ​​model to the device.

[0995] 6. Terminal: Displays the sent response on an interactive screen.

[0996] Specific examples

[0997] User: Type "Hello, Character A!" and submit.

[0998] Terminal: Sends a message to the server.

[0999] Server: Analyzes the message content and has the AI ​​model generate a response to the "hello" part.

[1000] Server: Generates a response saying "Hello, I'm Character A!" and sends it to the device.

[1001] Terminal: Display "Hello, I'm Character A!" on the dialogue screen.

[1002] Gathering feedback

[1003] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1004] 2. User: Select and submit a rating.

[1005] 3. Device: Sends the selected rating to the server.

[1006] 4. Server: Collects feedback and stores it in a reputation database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[1007] Tuning the model

[1008] 1. Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[1009] 2. Server: Identify areas for improvement and retrain the AI ​​model to improve performance if necessary.

[1010] 3. Server: Apply the improved model to the platform and provide it to users.

[1011] In this way, the system of the present invention allows for natural interaction with characters and continuously improves the performance of the model.

[1012] The processing flow will be explained below.

[1013] Step 1:

[1014] User: Launch the "Minkuri PF" application.

[1015] Terminal: Displays a login screen to the user and prompts them to enter their authentication information (username, password).

[1016] Step 2:

[1017] User: Enters authentication information and clicks the login button.

[1018] Terminal: Sends the entered authentication information to the server.

[1019] Step 3:

[1020] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[1021] Device: Receives a successful authentication message and displays the home screen.

[1022] Step 4:

[1023] Device: The home screen displays a list of characters that the user can choose from.

[1024] User: Select the character you want to interact with.

[1025] Step 5:

[1026] Device: Sends the selected character ID to the server.

[1027] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[1028] Step 6:

[1029] Terminal: Displays an interactive screen and prompts the user to enter a message.

[1030] User: Enter a message and press send.

[1031] Step 7:

[1032] Terminal: Sends the entered message to the server.

[1033] Server: Parses the message and extracts important keywords and context.

[1034] Step 8:

[1035] Server: Based on the analysis results, the AI ​​model generates an appropriate response.

[1036] Server: Generates and sends the response to the device.

[1037] Step 9:

[1038] Terminal: Displays the received response on the interactive screen.

[1039] User: Type your next message and continue the conversation.

[1040] Step 10:

[1041] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1042] User: Select and submit a rating.

[1043] Step 11:

[1044] Device: Sends the user-selected rating to the server.

[1045] Server: Receives ratings and stores them in a feedback database.

[1046] Step 12:

[1047] Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[1048] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance.

[1049] Step 13:

[1050] Server: Applying the improved AI model to the platform and providing it to users.

[1051] Example 1

[1052] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1053] Conventional dialogue systems struggle to provide consistent responses when users interact with specific characters. Furthermore, feedback collection and AI model tuning are often done manually, resulting in slow performance improvements. Furthermore, insufficient user authentication processes create security challenges.

[1054] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1055] In this invention, the server includes means for inputting user authentication information and starting a session if authentication is successful, means for the user to select a specific character, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for the AI ​​model to generate an appropriate response based on the analysis result, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback, thereby making it possible to provide consistent and advanced responses, rapidly improve the model based on feedback, and achieve strong user authentication security.

[1056] "User authentication information" is information used to identify a user and verify access rights.

[1057] A "session" is a unit for recording and tracking a series of operations performed while a user is accessing a system.

[1058] A "character" is a virtual agent that interacts with the user within a dialogue system.

[1059] An "AI model" is an algorithm or system that uses artificial intelligence techniques to generate appropriate responses to given inputs.

[1060] An "input message" is information in text, voice, or other form that a user sends to a dialogue system.

[1061] "Parsing" is the process of understanding a received input message and identifying its meaning or intent.

[1062] A "response" is a reply message generated by an AI model in response to an input message.

[1063] "Feedback" refers to the ratings and opinions that users provide about their experience using the system.

[1064] "Tuning" is the process of improving the performance of an AI model based on collected feedback.

[1065] This invention provides a platform for users to interact with specific characters. The system allows users to select a character, loads an AI model corresponding to the selected character, analyzes messages from the user, and the AI ​​model generates an appropriate response and provides that response to the user. Furthermore, the system has the function of collecting user feedback and tuning the AI ​​model based on that feedback.

[1066] The hardware used includes servers (e.g., Dell PowerEdge R740) and user-operated devices (e.g., smartphones, tablets).The software used includes application software (e.g., "Minkuri PF") and AI model software (e.g., GPT-3, BERT).

[1067] Application launch and authentication

[1068] The user launches the Minkuri PF application on their device. The launched application displays a login screen and prompts the user to enter authentication information (user name and password).

[1069] When the user enters authentication information and presses the login button, the device sends the authentication information to the server. The server verifies whether the authentication information is correct, and if authentication is successful, starts a session and instructs the device to display the home screen.

[1070] Character Selection

[1071] The device will display a list of selectable characters on the home screen.

[1072] The user selects the character they want to interact with, and the device sends the selected character's ID to the server, which loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[1073] Starting a conversation

[1074] The terminal displays an interactive screen and prompts the user to input a message.

[1075] The user inputs a message they wish to communicate and presses the send button, and the terminal sends the message to the server.

[1076] The server analyzes the received message and inputs it into the AI ​​model, which generates an appropriate response that the server then sends to the device.

[1077] The terminal displays the transmitted response on an interactive screen.

[1078] As a concrete example, consider a scenario where a user types and sends "Hello, Character A!" In this case, the server analyzes the message content and has the AI ​​model generate an appropriate response for the "Hello" part. The generated response is "Hello, I'm Character A!", which is sent to the device and displayed on the interactive screen.

[1079] Collecting feedback

[1080] After the interaction is completed, the terminal displays a screen asking the user to rate the interaction experience (e.g., five stars).

[1081] The user selects and submits a rating. The device then sends the selected rating to the server. The server collects the feedback and stores it in a rating database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[1082] Tuning the model

[1083] The server periodically analyzes the feedback data to evaluate the performance of the AI ​​model, finds areas for improvement, and retrains the AI ​​model to improve its performance as needed. The improved model is then applied to the platform and provided to users.

[1084] In this way, our system allows for natural interaction with characters and can continuously improve the model's performance based on feedback.

[1085] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1086] Step 1: Launch the Application and Authenticate

[1087] The user launches the application on the device and is presented with a login screen where they can enter their username and password.

[1088] The user enters authentication information (user name, password) and presses the login button.

[1089] The terminal sends the entered authentication information to the server. At this time, the authentication information is sent in encrypted form.

[1090] The server checks the authentication information against the database, and if it matches, it sends a response indicating successful authentication to the terminal. If it does not match, it sends a message indicating unsuccessful authentication.

[1091] Input: Username and Password

[1092] Output: Session ID if authentication is successful, error message if authentication fails

[1093] What it does: Send and verify credentials

[1094] Step 2: Character Selection

[1095] After successful authentication, the device will display a list of selectable characters on the home screen.

[1096] The user selects the character with which they want to interact.

[1097] The terminal transmits the selected character ID to the server.

[1098] Input: User's character selection

[1099] Output: Selected character ID

[1100] Action: Display and select a character from the list

[1101] Step 3: Loading the AI ​​model

[1102] The server loads the AI ​​model corresponding to the received character ID.

[1103] The server notifies the device that the AI ​​model has finished loading.

[1104] Input: Character ID

[1105] Output: Notification that the AI ​​model is ready to load

[1106] Action: Loading an AI model

[1107] Step 4: Start a conversation

[1108] The terminal displays an interactive screen and prompts the user to input a message.

[1109] The user inputs a message to be communicated and presses the send button.

[1110] The terminal transmits the input message to the server.

[1111] Input: User's message

[1112] Output: Message sent to server

[1113] Action: Type and send a message

[1114] Step 5: Parsing the message and generating a response

[1115] The server analyzes the received message and inputs its contents into the AI ​​model.

[1116] The server receives the response generated by the AI ​​model and sends it to the device.

[1117] Input: User's message

[1118] Output: Response from the AI ​​model

[1119] Action: Parse the message and generate a response

[1120] Step 6: Display the response

[1121] The terminal displays the response received from the server on an interactive screen.

[1122] Input: Response from the server

[1123] Output: Response display on the interactive screen

[1124] Action: Display response

[1125] Step 7: Gather feedback

[1126] After the interaction is completed, the terminal displays a screen requesting the user to evaluate the interaction experience.

[1127] The user selects and submits a rating.

[1128] The terminal transmits the selected rating to the server.

[1129] Input: User rating

[1130] Output: Send rating to server

[1131] Action: Enter and submit a rating

[1132] Step 8: Storing and analyzing feedback

[1133] The server stores the received feedback in a ratings database.

[1134] The server analyzes the feedback data and uses it to improve the performance of the AI ​​model.

[1135] Input: User feedback

[1136] Output: Save to database and analysis results

[1137] How it works: Saving and analyzing feedback

[1138] Step 9: Tune the AI ​​model

[1139] The server periodically analyzes the feedback data and retrains the AI ​​model as needed to improve its performance.

[1140] The server applies the improved model to the platform and provides it to the user.

[1141] Input: Feedback data

[1142] Output: Improved model

[1143] How it works: Retraining and applying AI models

[1144] The system's processing allows for natural interaction with characters and allows for continuous performance improvement.

[1145] (Application example 1)

[1146] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1147] In today's brick-and-mortar stores, many customers want an environment where they can easily ask questions about detailed product information and recommended products. However, with limited staff numbers, it is difficult to respond to all customer inquiries quickly and appropriately. There is also a lack of efficient ways to collect customer feedback and reflect it in service improvements. As a result, customer satisfaction may decline, which could have a negative impact on store sales. An effective way to solve this problem is needed.

[1148] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1149] In this invention, the server includes a means for a user to select a specific character, a means for loading an AI model corresponding to the selected character, a means for receiving and analyzing an input message from the user, and a means for applying the generated response to product introductions and inquiries in a physical store. This allows for automated dialogue with customers, enabling prompt and appropriate responses. Furthermore, tuning the AI ​​model based on collected feedback enables continuous improvement of the service.

[1150] "User" refers to any individual or corporation that uses this system.

[1151] "Character" refers to a virtual person or character that a user selects as a target for interaction.

[1152] "AI Model" refers to the artificial intelligence algorithms and data models that generate responses to selected characters.

[1153] "Input message" refers to text or voice information sent by a user to the system.

[1154] "Parsing" refers to understanding the content of an input message and processing it to generate an appropriate response.

[1155] "Response" refers to the reply or information provided to the user generated by the AI ​​model.

[1156] "Feedback" refers to the ratings and opinions provided by users about their interactive experiences.

[1157] "Tuning" refers to making adjustments to improve the performance and accuracy of an AI model based on collected feedback.

[1158] "Brick and mortar store" refers to a physical location for selling products.

[1159] "Product introduction" refers to explaining the features and benefits of a product to customers in a physical store.

[1160] "Inquiry response" refers to responding to customer questions and requests and providing appropriate information and services.

[1161] The present invention includes a system that provides a platform for users to interact with specific characters and effectively introduces products and responds to inquiries in physical stores.

[1162] The server first receives user authentication information and allows the user to log in to the system. After authentication, the user is asked to select from multiple characters and the corresponding AI model is loaded based on the selection. The server then receives the input message from the user, analyzes the message, and generates an appropriate response. The generated response is used in the physical store to introduce products and respond to inquiries. In addition, feedback provided by the user after the interaction is collected and used to tune the AI ​​model, thereby continuously improving the quality of service.

[1163] The main hardware used is the device used by the user, such as a smartphone or smart glasses, while the server side requires computing resources to run a high-performance AI model. This is achieved by using cloud computing services (e.g., Amazon Web Services or Google Cloud Platform). On the software side, Flask is used to provide an API, and data communication is often in JSON format.

[1164] Specifically, the user launches the app on their smartphone, logs in, and then selects the "Shopping Assist Character." An example of the prompt is shown below.

[1165] User: "What are the features of this product?"

[1166] AI: "This product incorporates the latest technology and combines functionality with design."

[1167] Through these interactions, customers can quickly and accurately obtain the information they are looking for. After the interaction is over, the user provides a star rating, which is sent as feedback to the server. The server then uses this feedback to retrain the AI ​​model and improve the overall system so that it can provide more appropriate responses in the next interaction.

[1168] By introducing this system, customer service in physical stores will be automated, and it is expected that customer satisfaction will improve and store operations will become more efficient.

[1169] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1170] Step 1: User selects a specific character

[1171] Input: The user launches the app and enters their login information (username, password).

[1172] Data processing: The device sends the entered login information to the server, which checks the authentication information against its database.

[1173] Output: If authentication is successful, a list of characters the user can choose from is displayed on the terminal.

[1174] Step 2: Load the AI ​​model corresponding to the selected character

[1175] Input: The user selects a character. For example, "Shopping Assist Character."

[1176] Data processing: The device sends the selected character ID to the server, which then loads the AI ​​model corresponding to the character ID into memory.

[1177] Output: Notifies that a character has been selected and displays the dialogue screen on the terminal.

[1178] Step 3: Receive and parse the user input message

[1179] Input: The user types a message into the interactive screen and sends it. For example, "What are the features of this product?"

[1180] Data processing: The device sends the input message to the server, which analyzes the message using techniques such as morphological analysis to extract meaning.

[1181] Output: Based on the analysis results, a prompt is generated for the AI ​​model, ready to generate an appropriate response.

[1182] Step 4: The AI ​​model generates an appropriate response

[1183] Input: Parsed message content (prompt sentence). For example, "User: What are the features of this product?"

[1184] Data processing: The server inputs the analysis results into the AI ​​model, and the generative AI model generates a response.

[1185] Output: The generated response, for example, "This product uses the latest technology and combines functionality and design."

[1186] Step 5: Providing the generated response to the user

[1187] Input: The response sentence generated by the AI ​​model.

[1188] Data processing: The server sends the generated response to the terminal, which then displays the received response on the interactive screen.

[1189] Output: A response to be displayed on the interactive screen. For example, "This product uses the latest technology and combines functionality and design."

[1190] Step 6: Gather user feedback

[1191] Input: A feedback request screen after the interaction is completed. The user can enter and submit an evaluation of their interaction experience.

[1192] Data processing: The device sends the entered rating to the server, which stores the rating in the rating database.

[1193] Output: The evaluation data is saved in a database.

[1194] Step 7: Tune the AI ​​model based on collected feedback

[1195] Input: Collected feedback data.

[1196] Data processing: The server analyzes the feedback data and evaluates the performance of the AI ​​model, retraining it as needed and tuning the model.

[1197] Output: A new, improved AI model that will enable a better response the next time you interact with it.

[1198] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1199] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[1200] Application launch and authentication

[1201] 1. User: Launch the "Minkuri PF" application.

[1202] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[1203] 3. User: After entering the authentication information, click the login button.

[1204] 4. Device: Sends authentication information to the server.

[1205] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[1206] Character Selection

[1207] 1. Device: Display a list of characters that the user can select from on the home screen.

[1208] 2. User: Select the character you want to interact with.

[1209] 3. Device: Send the selected character ID to the server.

[1210] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[1211] Dialogue initiation and emotion recognition

[1212] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[1213] 2. User: Enter a message and press the send button.

[1214] 3. Terminal: Sends the entered message to the server.

[1215] 4. Server: Receives the message and detects the user's emotion using the emotion engine.

[1216] 5. Server: Add emotional information to the message content analysis results and input it into the AI ​​model.

[1217] 6. Server: Based on the emotional information, the AI ​​model generates an appropriate response and sends it to the device.

[1218] 7. Terminal: Displays the received response on the interactive screen.

[1219] Specific examples

[1220] User: Type "I'm really tired today" and submit.

[1221] Terminal: Sends a message to the server.

[1222] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[1223] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[1224] Server: Generates and sends the response to the device.

[1225] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[1226] Gathering feedback

[1227] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1228] 2. User: Select and submit a rating.

[1229] 3. Device: Sends user ratings to the server.

[1230] 4. Server: Stores the ratings in a feedback database and then analyzes the data to improve the performance of the AI ​​model.

[1231] Tuning the model

[1232] 1. Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[1233] 2. Server: Identify areas for improvement and retrain the AI ​​model as needed to improve performance.

[1234] 3. Server: Apply the improved AI model to the platform and provide it to users.

[1235] In this way, the system of the present invention, combined with emotion recognition capabilities, can achieve more natural and personalized interactions, providing responses based on the user's emotions, improving the interaction experience and increasing user satisfaction.

[1236] The processing flow will be explained below.

[1237] Step 1:

[1238] User: Launch the "Minkuri PF" application.

[1239] Terminal: Displays a login screen and prompts the user to enter authentication information (username, password).

[1240] Step 2:

[1241] User: Enters authentication information and clicks the login button.

[1242] Terminal: Sends the entered authentication information to the server.

[1243] Step 3:

[1244] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[1245] Device: Receives a successful authentication message and displays the home screen.

[1246] Step 4:

[1247] Device: The home screen displays a list of characters that the user can choose from.

[1248] User: Select the character you want to interact with.

[1249] Step 5:

[1250] Device: Sends the selected character ID to the server.

[1251] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[1252] Step 6:

[1253] Terminal: Displays an interactive screen and prompts the user to enter a message.

[1254] User: Enter a message and press send.

[1255] Step 7:

[1256] Terminal: Sends the entered message to the server.

[1257] Server: Receives the message and detects the user's emotion using the emotion engine.

[1258] Step 8:

[1259] Server: The analysis results include the emotional information detected by the emotion engine along with the message content.

[1260] Step 9:

[1261] Server: The AI ​​model generates an appropriate response taking into account emotional information.

[1262] Server: Generates and sends the response to the device.

[1263] Step 10:

[1264] Terminal: Display the response on the interactive screen.

[1265] User: Continue the conversation by entering a next message if desired.

[1266] Specific examples of emotion recognition

[1267] User: Type "I'm really tired today" and submit.

[1268] Terminal: Sends a message to the server.

[1269] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[1270] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[1271] Server: Generates and sends the response to the device.

[1272] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[1273] Step 11:

[1274] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1275] User: Select and submit a rating.

[1276] Step 12:

[1277] Device: Sends user ratings to the server.

[1278] Server: Stores the evaluations in a feedback database, then analyzes the data and uses it to improve the performance of the AI ​​model.

[1279] Step 13:

[1280] Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[1281] Server: Retrains the AI ​​model as needed to improve its performance.

[1282] Server: Applying the improved AI model to the platform and providing it to users.

[1283] Example 2

[1284] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1285] Conventional dialogue systems often provide simple responses without understanding the user's emotions. This results in low-quality dialogue and makes it difficult to improve user satisfaction. Furthermore, they lack the means to effectively utilize feedback to improve the system, making it difficult to respond quickly to user needs.

[1286] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1287] In this invention, the server includes: a means for a user to select a specific character; a means for loading an AI model corresponding to the selected character; a means for receiving and analyzing an input message from the user; a means for the AI ​​model to generate an appropriate response based on the analysis result; a means for providing the generated response to the user; a means for collecting feedback provided by the user; a means for adjusting the AI ​​model based on the collected feedback; a means for analyzing the input message using an emotion engine to recognize the user's emotions; and a means for the AI ​​model to generate a response based on the emotion information. This enables the provision of higher-quality dialogue based on the user's emotions, thereby improving user satisfaction. Furthermore, the AI ​​model can be continuously adjusted and improved using feedback, enabling the system to quickly respond to user needs.

[1288] "User" refers to a human being who uses a dialogue system.

[1289] A "character" is a visual or conceptual representation of a person or object that is the subject of a dialogue, and is used to interact with the user in a dialogue system.

[1290] An "artificial intelligence model" refers to an algorithm that learns from large amounts of data to generate and analyze text, and specifically includes generative AI models (e.g., GPT-3).

[1291] An "input message" refers to information such as text or voice that a user sends to a dialogue system.

[1292] "Analysis" refers to the act of using natural language processing technology to understand the meaning and emotions of an input message and process the information.

[1293] "Response" refers to a reply or message to the user that the artificial intelligence model generates based on the analysis results.

[1294] An "emotion engine" refers to an algorithm or system for detecting and classifying a user's emotions from an input message.

[1295] "Feedback" refers to information such as evaluations and opinions that users provide to a dialogue system.

[1296] "Review Database" refers to a database for storing and managing feedback collected from users.

[1297] "Tuning" refers to the act of retraining or changing parameters to improve the performance or behavior of an AI model based on collected feedback and other data.

[1298] MODE FOR CARRYING OUT THE INVENTION

[1299] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[1300] First, the user launches an application on a device such as a smartphone, tablet, or PC. This application provides the interface necessary for the interactive system. Specific application names include "Minkuri PF."

[1301] When a user launches an application, the device displays a login screen and prompts for a username and password. The user enters the authentication information and presses the login button, and the device sends the authentication information to the server. The server verifies whether the received authentication information is correct, and if so, starts a session and instructs the device to display the home screen.

[1302] When the home screen is displayed, the device presents the user with a list of characters to choose from. The user selects the character they want to interact with, and the device sends the selected character's ID to the server. The server loads the AI ​​model (e.g., GPT-3) corresponding to the character ID and notifies the device that it is ready.

[1303] When a conversation begins, the device displays a dialogue screen and prompts the user to enter a message. The user enters a message and presses the send button, and the device sends the message to the server. The server analyzes the received message using an emotion engine to recognize the user's emotions. By analyzing specific keywords and sentences, emotions can be classified as "joy," "sadness," "anger," "fatigue," etc.

[1304] Based on the analysis results, the server provides emotion information to the generative AI model, which then generates an appropriate response. The response is sent from the server to the device, which then displays it on the interactive screen. For example, if a user types, "I'm very tired today," the emotion engine detects the keyword "tired" and recognizes the emotion as "fatigue." The generative AI model then generates a response such as, "Good work! Take a good rest," which is displayed on the device.

[1305] After the interaction is completed, the device displays a screen asking the user to rate the interaction experience, and the user enters and submits the rating. The device then sends the rating to the server, which stores it in a feedback database. The feedback is periodically analyzed and used to improve the AI ​​model.

[1306] The following are examples of prompt sentences:

[1307] The user types, "I'm very tired today." Recognize the user's emotions and generate an appropriate response.

[1308] To implement this system, the following hardware and software are required.

[1309] Hardware: devices such as smartphones, tablets, and computers

[1310] Software: conversational applications (e.g., Minkuri PF), server software, generative AI models (e.g., GPT-3), emotion recognition engines (e.g., NLTK, SpaCy)

[1311] In this way, the system of the present invention combines emotion recognition capabilities to enable more natural and personalized interactions, improving user satisfaction.

[1312] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1313] Step 1: Launch the Application and Authenticate

[1314] User: Launch the "Minkuri PF" application on a device such as a smartphone, tablet, or PC.

[1315] Terminal: Displays the login screen and prompts for a username and password. Receives the authentication information (username and password) entered by the user.

[1316] User: Enter your username and password and click the login button.

[1317] Terminal: Sends the entered information to the server. The input is (username, password) and the output is (sending authentication information).

[1318] Server: Verifies whether the received authentication information matches the data in the database. Verification involves performing a data check to compare the authentication information with the database. The output is (authentication result).

[1319] Server: Once authenticated, it starts a session and instructs the device to display the home screen. The output is (instruction to display home screen).

[1320] Step 2: Character Selection

[1321] Device: Displays the home screen and provides the user with a list of characters to choose from.

[1322] User: Select the character you want to interact with and tap on that character.

[1323] Terminal: Sends the selected character's ID to the server. Input is (selected character ID), output is (sent character ID).

[1324] Server: Based on the received character ID, load the corresponding AI model (e.g., GPT-3). The input is (character ID) and the output is (loaded AI model).

[1325] Server: Notify the terminal that loading is complete. The output is (notification of readiness).

[1326] Step 3: Initiating a dialogue and recognizing emotions

[1327] Terminal: Displays an interactive screen and prompts the user to enter a message.

[1328] User: Enter a message and press send.

[1329] Terminal: Sends the entered message to the server. The input is (user's message) and the output is (sent message).

[1330] Server: Analyzes received messages and detects user emotions using an emotion engine. The emotion engine uses natural language processing technology (e.g., NLTK, SpaCy) to analyze keywords in messages and assign emotion labels. The input is (message) and the output is (emotional information).

[1331] Server: The analysis results, including emotional information, are input into the generative AI model, which then generates an appropriate response. The input is (the analysis results, including emotional information), and the output is (the generated response).

[1332] Server: Sends the generated response to the terminal. The output is (sent response).

[1333] Terminal: Displays the received response on an interactive screen. The input is (the generated response) and the output is (the display of the response).

[1334] Step 4: Gather feedback

[1335] Terminal: After the interaction is completed, an evaluation screen is displayed, asking the user to rate their interaction experience.

[1336] User: Enter a rating and click the submit button.

[1337] Terminal: Sends the entered rating to the server. The input is (user rating) and the output is (sent rating).

[1338] Server: Stores the received ratings in a feedback database. The input is (user ratings) and the output is (storage of feedback data).

[1339] Step 5: Tune the model

[1340] Server: Periodically analyzes feedback data and emotion data to evaluate the performance of the AI ​​model. Performance evaluation uses statistical analysis and machine learning algorithms based on the collected data. The input is (feedback data, emotion data) and the output is (performance evaluation results).

[1341] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance. The input is (performance evaluation results) and the output is (an improved AI model).

[1342] Server: Apply the improved AI model to the platform and provide it to the user. The output is (application of the improved model).

[1343] (Application example 2)

[1344] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1345] Conventional self-driving vehicles lack the ability to recognize the emotional state of the driver and passengers in real time and provide appropriate responses or suggestions based on that information. This poses a problem in that it is not possible to reduce the mental and physical burden on the driver and passengers during long driving periods or stressful driving environments. Furthermore, there is a lack of a mechanism for utilizing feedback based on the user's emotions to improve the performance of AI models. This invention aims to solve these problems and improve the user experience in self-driving vehicles.

[1346] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1347] In this invention, the server includes a means for capturing a user's voice in real time and converting the voice to text, a means for analyzing the user's emotions using an emotion engine based on the analysis results, and a means for an AI model to generate an appropriate response based on the analysis results and emotion information. This enables appropriate responses and suggestions to be made according to the user's emotional state. Furthermore, by collecting feedback provided by users and continuously tuning the AI ​​model based on that feedback, the system's performance can be improved.

[1348] A "character" refers to a virtual entity that a user interacts with and that is visually displayed.

[1349] "AI model" refers to an algorithm for data analysis and response generation built using artificial intelligence.

[1350] "Emotion engine" refers to software or algorithms for analyzing messages received from a user and recognizing the user's emotions.

[1351] "Audio capture" refers to the process of collecting a user's voice using a microphone or the like.

[1352] "Text-to-text" refers to the process of converting audio data obtained through voice capture into written information.

[1353] "Response generation" refers to the process by which an AI model creates an appropriate response to a user based on analysis results and emotional information.

[1354] "Feedback" refers to evaluations and opinions provided by users regarding their experience using the system.

[1355] "Tuning" refers to the process of improving and adjusting the performance of an AI model based on collected feedback.

[1356] "Server" refers to a computer system that processes and analyzes various types of data.

[1357] "Real-time" refers to data processing and response generation occurring almost simultaneously with the passage of real time.

[1358] The present invention is a system for improving the user experience in an autonomous vehicle, and specific embodiments thereof are described below.

[1359] System Configuration

[1360] The system includes means for selecting a specific character with which a user will interact, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for analyzing the user's emotions using an emotion engine based on the analysis result, means for the AI ​​model to generate an appropriate response based on the analysis result and the emotion information, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback.

[1361] Hardware and Software

[1362] Hardware

[1363] Smart glasses or smartphone: Used for user voice input and feedback collection.

[1364] Server: Performs data processing and analysis, and hosts AI models.

[1365] software

[1366] Speech recognition engine (Google Speech Recognition API): Converts speech collected from a microphone into text.

[1367] Emotion analysis model (Hugging Face's Transformers library): Analyzes transcribed speech data to identify emotions.

[1368] Text-to-speech engine (pyttsx3): Provides responses generated by the AI ​​model as audio feedback to the user.

[1369] Data processing flow

[1370] First, the user uses smart glasses or a smartphone to input voice. This voice is converted into text through a speech recognition engine and sent to the server. The server then uses an emotion engine to analyze the emotion of the input text. Based on the analysis results and emotion information, the AI ​​model generates an appropriate response and provides feedback to the user as voice through a text-to-speech engine.

[1371] Specific examples

[1372] Input prompt example

[1373] "I'm very tired today."

[1374] "I'm tired from the long journey"

[1375] In response, the emotion engine identifies the emotion "fatigue," and the AI ​​model generates a response such as, "Thank you for your hard work. I'll play some relaxing music. Please take a short break at the next rest point." If the user provides feedback on this response, that feedback is collected and stored by the server and used to tune the AI ​​model in the future.

[1376] System Features

[1377] This system can provide appropriate responses and suggestions in real time according to the user's emotional state. For example, if the user feels tired while driving, it will automatically play relaxing music and guide them to appropriate rest stops. This helps reduce fatigue and stress caused by long driving hours.

[1378] The above is a specific embodiment of the present invention. Use of this system is expected to significantly improve the user experience in autonomous vehicles.

[1379] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1380] Step 1:

[1381] A user uses smart glasses or a smartphone to provide voice input.

[1382] Input: User's voice

[1383] Output: Audio data

[1384] Action: The user speaks to the device, saying something like "I'm very tired today."

[1385] Step 2:

[1386] The device uses a speech recognition engine (Google Speech Recognition API) to convert the voice data into text.

[1387] Input: Audio data

[1388] Output: Text data

[1389] How it works: The device's microphone captures audio and sends it to the Google Speech Recognition API, which converts it into text: "I'm very tired today."

[1390] Step 3:

[1391] The terminal transmits the text data to the server.

[1392] Input: Text data

[1393] Output: Request sent to server

[1394] Action: The device sends the converted text "I'm very tired today" to the server.

[1395] Step 4:

[1396] The server analyzes the emotions in the text data using an emotion engine (Hugging Face's Transformers library).

[1397] Input: Text data

[1398] Output: Emotional information (e.g., "fatigue")

[1399] How it works: The server inputs text data into the emotion engine and identifies the emotion "fatigue" from the keyword "tired."

[1400] Step 5:

[1401] The server uses an AI model to generate an appropriate response based on the analysis results and emotional information.

[1402] Input: Emotion information and text data

[1403] Output: Response text (e.g. "Good work! We'll play some relaxing music. Please take a short break at the next break point.")

[1404] How it works: Based on the emotional information "fatigue," the server uses an AI model to generate an appropriate response, which is then output as text.

[1405] Step 6:

[1406] The server sends the generated response text to the terminal.

[1407] Input: Response text

[1408] Output: Sending a request to the terminal

[1409] Behavior: The server generates a response "Good work! We'll play some relaxing music. Please take a short break at the next rest point." and sends it to the device.

[1410] Step 7:

[1411] The device uses a text-to-speech engine (pyttsx3) to audibly output the response text.

[1412] Input: Response text

[1413] Output: Audio data

[1414] How it works: The device inputs the response text into a text-to-speech engine and provides it as audio feedback to the user.

[1415] Step 8:

[1416] The user provides feedback on the response.

[1417] Input: User ratings and opinions

[1418] Output: Feedback data

[1419] How it works: The user enters a rating or opinion on the device.

[1420] Step 9:

[1421] The terminal transmits the feedback data to the server.

[1422] Input: Feedback data

[1423] Output: Request sent to server

[1424] Operation: The terminal transmits the entered evaluation data to the server.

[1425] Step 10:

[1426] The server stores the collected feedback data and uses it to tune the AI ​​model.

[1427] Input: Feedback data

[1428] Output: An improved AI model

[1429] How it works: The server stores the feedback data in a database and periodically analyzes the data to improve the performance of the AI ​​model.

[1430] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1431] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1432] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1433] [Fourth embodiment]

[1434] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1435] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1436] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1437] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1438] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1439] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1440] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1441] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1442] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1443] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1444] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1445] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1446] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1447] This invention provides a platform for users to interact with specific characters, and includes the following processes: the user selects a character, loads an AI model corresponding to the selected character, analyzes messages from the user, the AI ​​model generates an appropriate response, and provides that response to the user. Furthermore, the system has the function of collecting feedback from users and tuning the AI ​​model based on that feedback.

[1448] Application launch and authentication

[1449] 1. User: Launch the "Minkuri PF" application.

[1450] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[1451] 3. User: After entering the authentication information, click the login button.

[1452] 4. Device: Sends authentication information to the server.

[1453] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[1454] Character Selection

[1455] 1. Device: Display a list of selectable characters on the home screen.

[1456] 2. User: Select the character you want to interact with.

[1457] 3. Device: Send the selected character ID to the server.

[1458] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[1459] Starting a conversation

[1460] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[1461] 2. User: Enter the message you want to communicate and press the send button.

[1462] 3. Terminal: Sends a message to the server.

[1463] 4. Server: Analyzes the received message and inputs it into the AI ​​model.

[1464] 5. Server: Sends the appropriate response generated by the AI ​​model to the device.

[1465] 6. Terminal: Displays the sent response on an interactive screen.

[1466] Specific examples

[1467] User: Type "Hello, Character A!" and submit.

[1468] Terminal: Sends a message to the server.

[1469] Server: Analyzes the message content and has the AI ​​model generate a response to the "hello" part.

[1470] Server: Generates a response saying "Hello, I'm Character A!" and sends it to the device.

[1471] Terminal: Display "Hello, I'm Character A!" on the dialogue screen.

[1472] Gathering feedback

[1473] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1474] 2. User: Select and submit a rating.

[1475] 3. Device: Sends the selected rating to the server.

[1476] 4. Server: Collects feedback and stores it in a reputation database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[1477] Tuning the model

[1478] 1. Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[1479] 2. Server: Identify areas for improvement and retrain the AI ​​model to improve performance if necessary.

[1480] 3. Server: Apply the improved model to the platform and provide it to users.

[1481] In this way, the system of the present invention allows for natural interaction with characters and continuously improves the performance of the model.

[1482] The processing flow will be explained below.

[1483] Step 1:

[1484] User: Launch the "Minkuri PF" application.

[1485] Terminal: Displays a login screen to the user and prompts them to enter their authentication information (username, password).

[1486] Step 2:

[1487] User: Enters authentication information and clicks the login button.

[1488] Terminal: Sends the entered authentication information to the server.

[1489] Step 3:

[1490] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[1491] Device: Receives a successful authentication message and displays the home screen.

[1492] Step 4:

[1493] Device: The home screen displays a list of characters that the user can choose from.

[1494] User: Select the character you want to interact with.

[1495] Step 5:

[1496] Device: Sends the selected character ID to the server.

[1497] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[1498] Step 6:

[1499] Terminal: Displays an interactive screen and prompts the user to enter a message.

[1500] User: Enter a message and press send.

[1501] Step 7:

[1502] Terminal: Sends the entered message to the server.

[1503] Server: Parses the message and extracts important keywords and context.

[1504] Step 8:

[1505] Server: Based on the analysis results, the AI ​​model generates an appropriate response.

[1506] Server: Generates and sends the response to the device.

[1507] Step 9:

[1508] Terminal: Displays the received response on the interactive screen.

[1509] User: Type your next message and continue the conversation.

[1510] Step 10:

[1511] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1512] User: Select and submit a rating.

[1513] Step 11:

[1514] Device: Sends the user-selected rating to the server.

[1515] Server: Receives ratings and stores them in a feedback database.

[1516] Step 12:

[1517] Server: Periodically analyzes feedback data and evaluates the performance of the AI ​​model.

[1518] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance.

[1519] Step 13:

[1520] Server: Applying the improved AI model to the platform and providing it to users.

[1521] Example 1

[1522] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1523] Conventional dialogue systems struggle to provide consistent responses when users interact with specific characters. Furthermore, feedback collection and AI model tuning are often done manually, resulting in slow performance improvements. Furthermore, insufficient user authentication processes create security challenges.

[1524] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1525] In this invention, the server includes means for inputting user authentication information and starting a session if authentication is successful, means for the user to select a specific character, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for the AI ​​model to generate an appropriate response based on the analysis result, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback, thereby making it possible to provide consistent and advanced responses, rapidly improve the model based on feedback, and achieve strong user authentication security.

[1526] "User authentication information" is information used to identify a user and verify access rights.

[1527] A "session" is a unit for recording and tracking a series of operations performed while a user is accessing a system.

[1528] A "character" is a virtual agent that interacts with the user within a dialogue system.

[1529] An "AI model" is an algorithm or system that uses artificial intelligence techniques to generate appropriate responses to given inputs.

[1530] An "input message" is information in text, voice, or other form that a user sends to a dialogue system.

[1531] "Parsing" is the process of understanding a received input message and identifying its meaning or intent.

[1532] A "response" is a reply message generated by an AI model in response to an input message.

[1533] "Feedback" refers to the ratings and opinions that users provide about their experience using the system.

[1534] "Tuning" is the process of improving the performance of an AI model based on collected feedback.

[1535] This invention provides a platform for users to interact with specific characters. The system allows users to select a character, loads an AI model corresponding to the selected character, analyzes messages from the user, and the AI ​​model generates an appropriate response and provides that response to the user. Furthermore, the system has the function of collecting user feedback and tuning the AI ​​model based on that feedback.

[1536] The hardware used includes servers (e.g., Dell PowerEdge R740) and user-operated devices (e.g., smartphones, tablets).The software used includes application software (e.g., "Minkuri PF") and AI model software (e.g., GPT-3, BERT).

[1537] Application launch and authentication

[1538] The user launches the Minkuri PF application on their device. The launched application displays a login screen and prompts the user to enter authentication information (user name and password).

[1539] When the user enters authentication information and presses the login button, the device sends the authentication information to the server. The server verifies whether the authentication information is correct, and if authentication is successful, starts a session and instructs the device to display the home screen.

[1540] Character Selection

[1541] The device will display a list of selectable characters on the home screen.

[1542] The user selects the character they want to interact with, and the device sends the selected character's ID to the server, which loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[1543] Starting a conversation

[1544] The terminal displays an interactive screen and prompts the user to input a message.

[1545] The user inputs a message they wish to communicate and presses the send button, and the terminal sends the message to the server.

[1546] The server analyzes the received message and inputs it into the AI ​​model, which generates an appropriate response that the server then sends to the device.

[1547] The terminal displays the transmitted response on an interactive screen.

[1548] As a concrete example, consider a scenario where a user types and sends "Hello, Character A!" In this case, the server analyzes the message content and has the AI ​​model generate an appropriate response for the "Hello" part. The generated response is "Hello, I'm Character A!", which is sent to the device and displayed on the interactive screen.

[1549] Gathering feedback

[1550] After the interaction is completed, the terminal displays a screen asking the user to rate the interaction experience (e.g., five stars).

[1551] The user selects and submits a rating. The device then sends the selected rating to the server. The server collects the feedback and stores it in a rating database. The feedback data is then analyzed and used to improve the performance of the AI ​​model.

[1552] Tuning the model

[1553] The server periodically analyzes the feedback data to evaluate the performance of the AI ​​model, finds areas for improvement, and retrains the AI ​​model to improve its performance as needed. The improved model is then applied to the platform and provided to users.

[1554] In this way, our system allows for natural interaction with characters and can continuously improve the model's performance based on feedback.

[1555] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1556] Step 1: Launch the Application and Authenticate

[1557] The user launches the application on the device and is presented with a login screen where they can enter their username and password.

[1558] The user enters authentication information (user name, password) and presses the login button.

[1559] The terminal sends the entered authentication information to the server. At this time, the authentication information is sent in encrypted form.

[1560] The server checks the authentication information against the database, and if it matches, it sends a response indicating successful authentication to the terminal. If it does not match, it sends a message indicating unsuccessful authentication.

[1561] Input: Username and Password

[1562] Output: Session ID if authentication is successful, error message if authentication fails

[1563] What it does: Send and verify credentials

[1564] Step 2: Character Selection

[1565] After successful authentication, the device will display a list of selectable characters on the home screen.

[1566] The user selects the character with which they want to interact.

[1567] The terminal transmits the selected character ID to the server.

[1568] Input: User's character selection

[1569] Output: Selected character ID

[1570] Action: Display and select a character from the list

[1571] Step 3: Loading the AI ​​model

[1572] The server loads the AI ​​model corresponding to the received character ID.

[1573] The server notifies the device that the AI ​​model has finished loading.

[1574] Input: Character ID

[1575] Output: Notification that the AI ​​model is ready to load

[1576] Action: Loading an AI model

[1577] Step 4: Start a conversation

[1578] The terminal displays an interactive screen and prompts the user to input a message.

[1579] The user inputs a message to be communicated and presses the send button.

[1580] The terminal transmits the input message to the server.

[1581] Input: User's message

[1582] Output: Message sent to server

[1583] Action: Type and send a message

[1584] Step 5: Parsing the message and generating a response

[1585] The server analyzes the received message and inputs its contents into the AI ​​model.

[1586] The server receives the response generated by the AI ​​model and sends it to the device.

[1587] Input: User's message

[1588] Output: Response from the AI ​​model

[1589] Action: Parse the message and generate a response

[1590] Step 6: Display the response

[1591] The terminal displays the response received from the server on an interactive screen.

[1592] Input: Response from the server

[1593] Output: Response display on the interactive screen

[1594] Action: Display response

[1595] Step 7: Gather feedback

[1596] After the interaction is completed, the terminal displays a screen requesting the user to evaluate the interaction experience.

[1597] The user selects and submits a rating.

[1598] The terminal transmits the selected rating to the server.

[1599] Input: User rating

[1600] Output: Send rating to server

[1601] Action: Enter and submit a rating

[1602] Step 8: Storing and analyzing feedback

[1603] The server stores the received feedback in a ratings database.

[1604] The server analyzes the feedback data and uses it to improve the performance of the AI ​​model.

[1605] Input: User feedback

[1606] Output: Save to database and analysis results

[1607] How it works: Saving and analyzing feedback

[1608] Step 9: Tune the AI ​​model

[1609] The server periodically analyzes the feedback data and retrains the AI ​​model as needed to improve its performance.

[1610] The server applies the improved model to the platform and provides it to the user.

[1611] Input: Feedback data

[1612] Output: Improved model

[1613] How it works: Retraining and applying AI models

[1614] The system's processing allows for natural interaction with characters and allows for continuous performance improvement.

[1615] (Application example 1)

[1616] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1617] In today's brick-and-mortar stores, many customers want an environment where they can easily ask questions about detailed product information and recommended products. However, with limited staff numbers, it is difficult to respond to all customer inquiries quickly and appropriately. There is also a lack of efficient ways to collect customer feedback and reflect it in service improvements. As a result, customer satisfaction may decline, which could have a negative impact on store sales. An effective way to solve this problem is needed.

[1618] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1619] In this invention, the server includes a means for a user to select a specific character, a means for loading an AI model corresponding to the selected character, a means for receiving and analyzing an input message from the user, and a means for applying the generated response to product introductions and inquiries in a physical store. This allows for automated dialogue with customers, enabling prompt and appropriate responses. Furthermore, tuning the AI ​​model based on collected feedback enables continuous improvement of the service.

[1620] "User" refers to any individual or corporation that uses this system.

[1621] "Character" refers to a virtual person or character that a user selects as a target for interaction.

[1622] "AI Model" refers to the artificial intelligence algorithms and data models that generate responses to selected characters.

[1623] "Input message" refers to text or voice information sent by a user to the system.

[1624] "Parsing" refers to understanding the content of an input message and processing it to generate an appropriate response.

[1625] "Response" refers to the reply or information provided to the user generated by the AI ​​model.

[1626] "Feedback" refers to the ratings and opinions provided by users about their interactive experiences.

[1627] "Tuning" refers to making adjustments to improve the performance and accuracy of an AI model based on collected feedback.

[1628] "Brick and mortar store" refers to a physical location for selling products.

[1629] "Product introduction" refers to explaining the features and benefits of a product to customers in a physical store.

[1630] "Inquiry response" refers to responding to customer questions and requests and providing appropriate information and services.

[1631] The present invention includes a system that provides a platform for users to interact with specific characters and effectively introduces products and responds to inquiries in physical stores.

[1632] The server first receives user authentication information and allows the user to log in to the system. After authentication, the user is asked to select from multiple characters and the corresponding AI model is loaded based on the selection. The server then receives the input message from the user, analyzes the message, and generates an appropriate response. The generated response is used in the physical store to introduce products and respond to inquiries. In addition, feedback provided by the user after the interaction is collected and used to tune the AI ​​model, thereby continuously improving the quality of service.

[1633] The main hardware used is the device used by the user, such as a smartphone or smart glasses, while the server side requires computing resources to run a high-performance AI model. This is achieved by using cloud computing services (e.g., Amazon Web Services or Google Cloud Platform). On the software side, Flask is used to provide an API, and data communication is often in JSON format.

[1634] Specifically, the user launches the app on their smartphone, logs in, and then selects the "Shopping Assist Character." An example of the prompt is shown below.

[1635] User: "What are the features of this product?"

[1636] AI: "This product incorporates the latest technology and combines functionality with design."

[1637] Through these interactions, customers can quickly and accurately obtain the information they are looking for. After the interaction is over, the user provides a star rating, which is sent as feedback to the server. The server then uses this feedback to retrain the AI ​​model and improve the overall system so that it can provide more appropriate responses in the next interaction.

[1638] By introducing this system, customer service in physical stores will be automated, and it is expected that customer satisfaction will improve and store operations will become more efficient.

[1639] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1640] Step 1: User selects a specific character

[1641] Input: The user launches the app and enters their login information (username, password).

[1642] Data processing: The device sends the entered login information to the server, which checks the authentication information against its database.

[1643] Output: If authentication is successful, a list of characters the user can choose from is displayed on the terminal.

[1644] Step 2: Load the AI ​​model corresponding to the selected character

[1645] Input: The user selects a character. For example, "Shopping Assist Character."

[1646] Data processing: The device sends the selected character ID to the server, which then loads the AI ​​model corresponding to the character ID into memory.

[1647] Output: Notifies that a character has been selected and displays the dialogue screen on the terminal.

[1648] Step 3: Receive and parse the user input message

[1649] Input: The user types a message into the interactive screen and sends it. For example, "What are the features of this product?"

[1650] Data processing: The device sends the input message to the server, which analyzes the message using techniques such as morphological analysis to extract meaning.

[1651] Output: Based on the analysis results, a prompt is generated for the AI ​​model, ready to generate an appropriate response.

[1652] Step 4: The AI ​​model generates an appropriate response

[1653] Input: Parsed message content (prompt sentence). For example, "User: What are the features of this product?"

[1654] Data processing: The server inputs the analysis results into the AI ​​model, and the generative AI model generates a response.

[1655] Output: The generated response, for example, "This product uses the latest technology and combines functionality and design."

[1656] Step 5: Providing the generated response to the user

[1657] Input: The response sentence generated by the AI ​​model.

[1658] Data processing: The server sends the generated response to the terminal, which then displays the received response on the interactive screen.

[1659] Output: A response to be displayed on the interactive screen. For example, "This product uses the latest technology and combines functionality and design."

[1660] Step 6: Gather user feedback

[1661] Input: A feedback request screen after the interaction is completed. The user can enter and submit an evaluation of their interaction experience.

[1662] Data processing: The device sends the entered rating to the server, which stores the rating in the rating database.

[1663] Output: The evaluation data is saved in a database.

[1664] Step 7: Tune the AI ​​model based on collected feedback

[1665] Input: Collected feedback data.

[1666] Data processing: The server analyzes the feedback data and evaluates the performance of the AI ​​model, retraining it as needed and tuning the model.

[1667] Output: A new, improved AI model that will enable a better response the next time you interact with it.

[1668] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1669] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[1670] Application launch and authentication

[1671] 1. User: Launch the "Minkuri PF" application.

[1672] 2. Terminal: Displays the login screen and prompts the user to enter authentication information (username, password).

[1673] 3. User: After entering the authentication information, click the login button.

[1674] 4. Device: Sends authentication information to the server.

[1675] 5. Server: Verifies that the authentication information is correct, and if authentication is successful, starts a session and instructs the home screen to be displayed.

[1676] Character Selection

[1677] 1. Device: Display a list of characters that the user can select from on the home screen.

[1678] 2. User: Select the character you want to interact with.

[1679] 3. Device: Send the selected character ID to the server.

[1680] 4. Server: Loads the AI ​​model corresponding to the character ID and notifies the device that it is ready.

[1681] Dialogue initiation and emotion recognition

[1682] 1. Terminal: Displays an interactive screen and prompts the user to enter a message.

[1683] 2. User: Enter a message and press the send button.

[1684] 3. Terminal: Sends the entered message to the server.

[1685] 4. Server: Receives the message and detects the user's emotion using the emotion engine.

[1686] 5. Server: Add emotional information to the message content analysis results and input it into the AI ​​model.

[1687] 6. Server: Based on the emotional information, the AI ​​model generates an appropriate response and sends it to the device.

[1688] 7. Terminal: Displays the received response on the interactive screen.

[1689] Specific examples

[1690] User: Type "I'm really tired today" and submit.

[1691] Terminal: Sends a message to the server.

[1692] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[1693] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[1694] Server: Generates and sends the response to the device.

[1695] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[1696] Gathering feedback

[1697] 1. Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1698] 2. User: Select and submit a rating.

[1699] 3. Device: Sends user ratings to the server.

[1700] 4. Server: Stores the ratings in a feedback database and then analyzes the data to improve the performance of the AI ​​model.

[1701] Tuning the model

[1702] 1. Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[1703] 2. Server: Identify areas for improvement and retrain the AI ​​model as needed to improve performance.

[1704] 3. Server: Apply the improved AI model to the platform and provide it to users.

[1705] In this way, the system of the present invention, combined with emotion recognition capabilities, can achieve more natural and personalized interactions, providing responses based on the user's emotions, improving the interaction experience and increasing user satisfaction.

[1706] The processing flow will be explained below.

[1707] Step 1:

[1708] User: Launch the "Minkuri PF" application.

[1709] Terminal: Displays a login screen and prompts the user to enter authentication information (username, password).

[1710] Step 2:

[1711] User: Enters authentication information and clicks the login button.

[1712] Terminal: Sends the entered authentication information to the server.

[1713] Step 3:

[1714] Server: Validates the authentication information and starts the session if it is correct, otherwise sends an error message to the terminal.

[1715] Device: Receives a successful authentication message and displays the home screen.

[1716] Step 4:

[1717] Device: The home screen displays a list of characters that the user can choose from.

[1718] User: Select the character you want to interact with.

[1719] Step 5:

[1720] Device: Sends the selected character ID to the server.

[1721] Server: Loads the AI ​​model corresponding to the character ID and notifies the device that the model is ready.

[1722] Step 6:

[1723] Terminal: Displays an interactive screen and prompts the user to enter a message.

[1724] User: Enter a message and press send.

[1725] Step 7:

[1726] Terminal: Sends the entered message to the server.

[1727] Server: Receives the message and detects the user's emotion using the emotion engine.

[1728] Step 8:

[1729] Server: The analysis results include the emotional information detected by the emotion engine along with the message content.

[1730] Step 9:

[1731] Server: The AI ​​model generates an appropriate response taking into account emotional information.

[1732] Server: Generates and sends the response to the device.

[1733] Step 10:

[1734] Terminal: Display the response on the interactive screen.

[1735] User: Continue the conversation by entering a next message if desired.

[1736] Specific examples of emotion recognition

[1737] User: Type "I'm really tired today" and submit.

[1738] Terminal: Sends a message to the server.

[1739] Server: The emotion engine detects the keyword "tired" and identifies the emotion as "fatigue."

[1740] Server: Emotional information is added to the message analysis results, and the AI ​​model generates a response such as "Good work! Take a good rest."

[1741] Server: Generates and sends the response to the device.

[1742] Terminal: Display "Good work! Have a good rest" on the dialogue screen.

[1743] Step 11:

[1744] Terminal: After the interaction is over, a screen is displayed asking the user to rate their interaction experience (e.g., 5 stars).

[1745] User: Select and submit a rating.

[1746] Step 12:

[1747] Device: Sends user ratings to the server.

[1748] Server: Stores the evaluations in a feedback database, then analyzes the data and uses it to improve the performance of the AI ​​model.

[1749] Step 13:

[1750] Server: Periodically analyzes feedback and sentiment data to evaluate the performance of the AI ​​model.

[1751] Server: Retrains the AI ​​model as needed to improve its performance.

[1752] Server: Applying the improved AI model to the platform and providing it to users.

[1753] Example 2

[1754] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1755] Conventional dialogue systems often provide simple responses without understanding the user's emotions. This results in low-quality dialogue and makes it difficult to improve user satisfaction. Furthermore, they lack the means to effectively utilize feedback to improve the system, making it difficult to respond quickly to user needs.

[1756] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1757] In this invention, the server includes: a means for a user to select a specific character; a means for loading an AI model corresponding to the selected character; a means for receiving and analyzing an input message from the user; a means for the AI ​​model to generate an appropriate response based on the analysis result; a means for providing the generated response to the user; a means for collecting feedback provided by the user; a means for adjusting the AI ​​model based on the collected feedback; a means for analyzing the input message using an emotion engine to recognize the user's emotions; and a means for the AI ​​model to generate a response based on the emotion information. This enables the provision of higher-quality dialogue based on the user's emotions, thereby improving user satisfaction. Furthermore, the AI ​​model can be continuously adjusted and improved using feedback, enabling the system to quickly respond to user needs.

[1758] "User" refers to a human being who uses a dialogue system.

[1759] A "character" is a visual or conceptual representation of a person or object that is the subject of a dialogue, and is used to interact with the user in a dialogue system.

[1760] An "artificial intelligence model" refers to an algorithm that learns from large amounts of data to generate and analyze text, and specifically includes generative AI models (e.g., GPT-3).

[1761] An "input message" refers to information such as text or voice that a user sends to a dialogue system.

[1762] "Analysis" refers to the act of using natural language processing technology to understand the meaning and emotions of an input message and process the information.

[1763] "Response" refers to a reply or message to the user that the artificial intelligence model generates based on the analysis results.

[1764] An "emotion engine" refers to an algorithm or system for detecting and classifying a user's emotions from an input message.

[1765] "Feedback" refers to information such as evaluations and opinions that users provide to a dialogue system.

[1766] "Review Database" refers to a database for storing and managing feedback collected from users.

[1767] "Tuning" refers to the act of retraining or changing parameters to improve the performance or behavior of an AI model based on collected feedback and other data.

[1768] MODE FOR CARRYING OUT THE INVENTION

[1769] The present invention provides a platform for users to interact with specific characters, and combines it with an emotion engine that recognizes the user's emotions during the interaction. The following is a specific embodiment of the system of the present invention.

[1770] First, the user launches an application on a device such as a smartphone, tablet, or PC. This application provides the interface necessary for the interactive system. Specific application names include "Minkuri PF."

[1771] When a user launches an application, the device displays a login screen and prompts for a username and password. The user enters the authentication information and presses the login button, and the device sends the authentication information to the server. The server verifies whether the received authentication information is correct, and if so, starts a session and instructs the device to display the home screen.

[1772] When the home screen is displayed, the device presents the user with a list of characters to choose from. The user selects the character they want to interact with, and the device sends the selected character's ID to the server. The server loads the AI ​​model (e.g., GPT-3) corresponding to the character ID and notifies the device that it is ready.

[1773] When a conversation begins, the device displays a dialogue screen and prompts the user to enter a message. The user enters a message and presses the send button, and the device sends the message to the server. The server analyzes the received message using an emotion engine to recognize the user's emotions. By analyzing specific keywords and sentences, emotions can be classified as "joy," "sadness," "anger," "fatigue," etc.

[1774] Based on the analysis results, the server provides emotion information to the generative AI model, which then generates an appropriate response. The response is sent from the server to the device, which then displays it on the interactive screen. For example, if a user types, "I'm very tired today," the emotion engine detects the keyword "tired" and recognizes the emotion as "fatigue." The generative AI model then generates a response such as, "Good work! Take a good rest," which is displayed on the device.

[1775] After the interaction is completed, the device displays a screen asking the user to rate the interaction experience, and the user enters and submits the rating. The device then sends the rating to the server, which stores it in a feedback database. The feedback is periodically analyzed and used to improve the AI ​​model.

[1776] The following are examples of prompt sentences:

[1777] The user types, "I'm very tired today." Recognize the user's emotions and generate an appropriate response.

[1778] To implement this system, the following hardware and software are required.

[1779] Hardware: devices such as smartphones, tablets, and computers

[1780] Software: conversational applications (e.g., Minkuri PF), server software, generative AI models (e.g., GPT-3), emotion recognition engines (e.g., NLTK, SpaCy)

[1781] In this way, the system of the present invention combines emotion recognition capabilities to enable more natural and personalized interactions, improving user satisfaction.

[1782] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1783] Step 1: Launch the Application and Authenticate

[1784] User: Launch the "Minkuri PF" application on a device such as a smartphone, tablet, or PC.

[1785] Terminal: Displays the login screen and prompts for a username and password. Receives the authentication information (username and password) entered by the user.

[1786] User: Enter your username and password and click the login button.

[1787] Terminal: Sends the entered information to the server. The input is (username, password) and the output is (sending authentication information).

[1788] Server: Verifies whether the received authentication information matches the data in the database. Verification involves performing a data check to compare the authentication information with the database. The output is (authentication result).

[1789] Server: Once authenticated, it starts a session and instructs the device to display the home screen. The output is (instruction to display home screen).

[1790] Step 2: Character Selection

[1791] Device: Displays the home screen and provides the user with a list of characters to choose from.

[1792] User: Select the character you want to interact with and tap on that character.

[1793] Terminal: Sends the selected character's ID to the server. Input is (selected character ID), output is (sent character ID).

[1794] Server: Based on the received character ID, load the corresponding AI model (e.g., GPT-3). The input is (character ID) and the output is (loaded AI model).

[1795] Server: Notify the terminal that loading is complete. The output is (notification of readiness).

[1796] Step 3: Initiating a dialogue and recognizing emotions

[1797] Terminal: Displays an interactive screen and prompts the user to enter a message.

[1798] User: Enter a message and press send.

[1799] Terminal: Sends the entered message to the server. The input is (user's message) and the output is (sent message).

[1800] Server: Analyzes received messages and detects user emotions using an emotion engine. The emotion engine uses natural language processing technology (e.g., NLTK, SpaCy) to analyze keywords in messages and assign emotion labels. The input is (message) and the output is (emotional information).

[1801] Server: The analysis results, including emotional information, are input into the generative AI model, which then generates an appropriate response. The input is (the analysis results, including emotional information), and the output is (the generated response).

[1802] Server: Sends the generated response to the terminal. The output is (sent response).

[1803] Terminal: Displays the received response on an interactive screen. The input is (the generated response) and the output is (the display of the response).

[1804] Step 4: Gather feedback

[1805] Terminal: After the interaction is completed, an evaluation screen is displayed, asking the user to rate their interaction experience.

[1806] User: Enter a rating and click the submit button.

[1807] Terminal: Sends the entered rating to the server. The input is (user rating) and the output is (sent rating).

[1808] Server: Stores the received ratings in a feedback database. The input is (user ratings) and the output is (storage of feedback data).

[1809] Step 5: Tune the model

[1810] Server: Periodically analyzes feedback data and emotion data to evaluate the performance of the AI ​​model. Performance evaluation uses statistical analysis and machine learning algorithms based on the collected data. The input is (feedback data, emotion data) and the output is (performance evaluation results).

[1811] Server: Identifies areas for improvement and retrains the AI ​​model as needed to improve performance. The input is (performance evaluation results) and the output is (an improved AI model).

[1812] Server: Apply the improved AI model to the platform and provide it to the user. The output is (application of the improved model).

[1813] (Application example 2)

[1814] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1815] Conventional self-driving vehicles lack the ability to recognize the emotional state of the driver and passengers in real time and provide appropriate responses or suggestions based on that information. This poses a problem in that it is not possible to reduce the mental and physical burden on the driver and passengers during long driving periods or stressful driving environments. Furthermore, there is a lack of a mechanism for utilizing feedback based on the user's emotions to improve the performance of AI models. This invention aims to solve these problems and improve the user experience in self-driving vehicles.

[1816] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1817] In this invention, the server includes a means for capturing a user's voice in real time and converting the voice to text, a means for analyzing the user's emotions using an emotion engine based on the analysis results, and a means for an AI model to generate an appropriate response based on the analysis results and emotion information. This enables appropriate responses and suggestions to be made according to the user's emotional state. Furthermore, by collecting feedback provided by users and continuously tuning the AI ​​model based on that feedback, the system's performance can be improved.

[1818] A "character" refers to a virtual entity that a user interacts with and that is visually displayed.

[1819] "AI model" refers to an algorithm for data analysis and response generation built using artificial intelligence.

[1820] "Emotion engine" refers to software or algorithms for analyzing messages received from a user and recognizing the user's emotions.

[1821] "Audio capture" refers to the process of collecting a user's voice using a microphone or the like.

[1822] "Text-to-text" refers to the process of converting audio data obtained through voice capture into written information.

[1823] "Response generation" refers to the process by which an AI model creates an appropriate response to a user based on analysis results and emotional information.

[1824] "Feedback" refers to evaluations and opinions provided by users regarding their experience using the system.

[1825] "Tuning" refers to the process of improving and adjusting the performance of an AI model based on collected feedback.

[1826] "Server" refers to a computer system that processes and analyzes various types of data.

[1827] "Real-time" refers to data processing and response generation occurring almost simultaneously with the passage of real time.

[1828] The present invention is a system for improving the user experience in an autonomous vehicle, and specific embodiments thereof are described below.

[1829] System Configuration

[1830] The system includes means for selecting a specific character with which a user will interact, means for loading an AI model corresponding to the selected character, means for receiving and analyzing an input message from the user, means for analyzing the user's emotions using an emotion engine based on the analysis result, means for the AI ​​model to generate an appropriate response based on the analysis result and the emotion information, means for providing the generated response to the user, means for collecting feedback provided by the user, and means for tuning the AI ​​model based on the collected feedback.

[1831] Hardware and Software

[1832] Hardware

[1833] Smart glasses or smartphone: Used for user voice input and feedback collection.

[1834] Server: Performs data processing and analysis, and hosts AI models.

[1835] software

[1836] Speech recognition engine (Google Speech Recognition API): Converts speech collected from a microphone into text.

[1837] Emotion analysis model (Hugging Face's Transformers library): Analyzes transcribed speech data to identify emotions.

[1838] Text-to-speech engine (pyttsx3): Provides responses generated by the AI ​​model as audio feedback to the user.

[1839] Data processing flow

[1840] First, the user uses smart glasses or a smartphone to input voice. This voice is converted into text through a speech recognition engine and sent to the server. The server then uses an emotion engine to analyze the emotion of the input text. Based on the analysis results and emotion information, the AI ​​model generates an appropriate response and provides feedback to the user as voice through a text-to-speech engine.

[1841] Specific examples

[1842] Input prompt statement example

[1843] "I'm very tired today."

[1844] "I'm tired from the long journey"

[1845] In response, the emotion engine identifies the emotion "fatigue," and the AI ​​model generates a response such as, "Thank you for your hard work. I'll play some relaxing music. Please take a short break at the next rest point." If the user provides feedback on this response, that feedback is collected and stored by the server and used to tune the AI ​​model in the future.

[1846] System Features

[1847] This system can provide appropriate responses and suggestions in real time according to the user's emotional state. For example, if the user feels tired while driving, it will automatically play relaxing music and guide them to appropriate rest stops. This helps reduce fatigue and stress caused by long driving hours.

[1848] The above is a specific embodiment of the present invention. Use of this system is expected to significantly improve the user experience in autonomous vehicles.

[1849] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1850] Step 1:

[1851] A user uses smart glasses or a smartphone to provide voice input.

[1852] Input: User's voice

[1853] Output: Audio data

[1854] Action: The user speaks to the device, saying something like "I'm very tired today."

[1855] Step 2:

[1856] The device uses a speech recognition engine (Google Speech Recognition API) to convert the voice data into text.

[1857] Input: Audio data

[1858] Output: Text data

[1859] How it works: The device's microphone captures audio and sends it to the Google Speech Recognition API, which converts it into text: "I'm very tired today."

[1860] Step 3:

[1861] The terminal transmits the text data to the server.

[1862] Input: Text data

[1863] Output: Request sent to server

[1864] Action: The device sends the converted text "I'm very tired today" to the server.

[1865] Step 4:

[1866] The server analyzes the emotions in the text data using an emotion engine (Hugging Face's Transformers library).

[1867] Input: Text data

[1868] Output: Emotional information (e.g., "fatigue")

[1869] How it works: The server inputs text data into the emotion engine and identifies the emotion "fatigue" from the keyword "tired."

[1870] Step 5:

[1871] The server uses an AI model to generate an appropriate response based on the analysis results and emotional information.

[1872] Input: Emotion information and text data

[1873] Output: Response text (e.g. "Good work! We'll play some relaxing music. Please take a short break at the next break point.")

[1874] How it works: Based on the emotional information "fatigue," the server uses an AI model to generate an appropriate response, which is then output as text.

[1875] Step 6:

[1876] The server sends the generated response text to the terminal.

[1877] Input: Response text

[1878] Output: Sending a request to the terminal

[1879] Behavior: The server generates a response "Good work! We'll play some relaxing music. Please take a short break at the next rest point." and sends it to the device.

[1880] Step 7:

[1881] The device uses a text-to-speech engine (pyttsx3) to audibly output the response text.

[1882] Input: Response text

[1883] Output: Audio data

[1884] How it works: The device inputs the response text into a text-to-speech engine and provides it as audio feedback to the user.

[1885] Step 8:

[1886] The user provides feedback on the response.

[1887] Input: User ratings and opinions

[1888] Output: Feedback data

[1889] How it works: The user enters a rating or opinion on the device.

[1890] Step 9:

[1891] The terminal transmits the feedback data to the server.

[1892] Input: Feedback data

[1893] Output: Request sent to server

[1894] Operation: The terminal transmits the entered evaluation data to the server.

[1895] Step 10:

[1896] The server stores the collected feedback data and uses it to tune the AI ​​model.

[1897] Input: Feedback data

[1898] Output: An improved AI model

[1899] How it works: The server stores the feedback data in a database and periodically analyzes the data to improve the performance of the AI ​​model.

[1900] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1901] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1902] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1903] The emotion identification model 59 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 an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1904] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1905] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1906] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1907] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1908] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs 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 a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1909] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1910] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1911] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1912] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1913] 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.

[1914] It is not necessary to store all 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 all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1915] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1916] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with 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). Also, the hardware resource that executes the specific processing may be a single processor.

[1917] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1918] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1919] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1920] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1921] The following is further disclosed regarding the above embodiment.

[1922] (Claim 1)

[1923] a means for a user to select a particular character;

[1924] means for loading an AI model corresponding to the selected character;

[1925] means for receiving and parsing an input message from a user;

[1926] A means for the AI ​​model to generate an appropriate response based on the analysis results; and

[1927] means for providing the generated response to a user;

[1928] a means for collecting user-provided feedback;

[1929] A means of tuning the AI ​​model based on the collected feedback; and

[1930] A system including:

[1931] (Claim 2)

[1932] 10. The system of claim 1, wherein the means for collecting feedback includes means for storing user ratings in a ratings database and for analyzing the data.

[1933] (Claim 3)

[1934] 2. The system according to claim 1, wherein the character with which the user interacts is visually displayed on a terminal operated by the user.

[1935] "Example 1"

[1936] (Claim 1)

[1937] a means for a user to select a particular character;

[1938] means for loading an AI model corresponding to the selected character;

[1939] means for receiving and parsing an input message from a user;

[1940] A means for the AI ​​model to generate an appropriate response based on the analysis results; and

[1941] means for providing the generated response to a user;

[1942] a means for collecting user-provided feedback;

[1943] A means of tuning the AI ​​model based on the collected feedback; and

[1944] a means for entering user authentication information and initiating a session if authentication is successful;

[1945] A system including:

[1946] (Claim 2)

[1947] 10. The system of claim 1, wherein the means for collecting feedback includes means for storing user ratings in a ratings database and for analyzing the data.

[1948] (Claim 3)

[1949] 2. The system according to claim 1, wherein the character with which the user interacts is visually displayed on a terminal operated by the user.

[1950] "Application Example 1"

[1951] (Claim 1)

[1952] a means for a user to select a particular character;

[1953] means for loading an AI model corresponding to the selected character;

[1954] means for receiving and parsing an input message from a user;

[1955] A means for the AI ​​model to generate an appropriate response based on the analysis results; and

[1956] means for providing the generated response to a user;

[1957] a means for collecting user-provided feedback;

[1958] A means of tuning the AI ​​model based on the collected feedback; and

[1959] A system that includes a means for applying the generated response content to product introductions and inquiry responses in physical stores.

[1960] (Claim 2)

[1961] 10. The system of claim 1, wherein the means for collecting feedback includes means for storing user ratings in a ratings database and for analyzing the data.

[1962] (Claim 3)

[1963] 2. The system according to claim 1, wherein the character with which the user interacts is visually displayed on a terminal operated by the user.

[1964] "Example 2: Combining Emotion Engines"

[1965] (Claim 1)

[1966] a means for a user to select a particular character;

[1967] means for loading an artificial intelligence model corresponding to the selected character;

[1968] means for receiving and parsing an input message from a user;

[1969] a means for the artificial intelligence model to generate an appropriate response based on the analysis results;

[1970] means for providing the generated response to a user;

[1971] a means for collecting user-provided feedback;

[1972] a means for adjusting the artificial intelligence model based on the collected feedback; and

[1973] means for analyzing the input message with an emotion engine to recognize the emotion of the user;

[1974] a means for the artificial intelligence model to generate a response based on the emotional information;

[1975] A system including:

[1976] (Claim 2)

[1977] 10. The system of claim 1, wherein the means for collecting feedback includes means for storing user ratings in a ratings database and for analyzing the data.

[1978] (Claim 3)

[1979] 2. The system according to claim 1, wherein the character with which the user interacts is visually displayed on a terminal operated by the user.

[1980] "Application example 2 when combining emotion engines"

[1981] (Claim 1)

[1982] a means for a user to select a particular character;

[1983] means for loading an AI model corresponding to the selected character;

[1984] means for receiving and parsing an input message from a user;

[1985] means for analyzing the user's emotions using an emotion engine based on the analysis results;

[1986] A means for the AI ​​model to generate an appropriate response based on the analysis results and emotional information; and

[1987] means for providing the generated response to a user;

[1988] a means for capturing a user's voice in real time and converting the voice to text;

[1989] a means for collecting user-provided feedback;

[1990] A means of tuning the AI ​​model based on the collected feedback; and

[1991] A system including:

[1992] (Claim 2)

[1993] 10. The system of claim 1, wherein the means for collecting feedback includes means for storing user ratings in a ratings database and for analyzing the data.

[1994] (Claim 3)

[1995] 2. The system according to claim 1, wherein the character with which the user interacts is visually displayed on a terminal operated by the user. [Explanation of symbols]

[1996] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for a user to select a particular character; means for loading an AI model corresponding to the selected character; means for receiving and parsing an input message from a user; A means for the AI ​​model to generate an appropriate response based on the analysis results; and means for providing the generated response to a user; a means for collecting user-provided feedback; A means of tuning the AI ​​model based on the collected feedback; and A system including:

2. 10. The system of claim 1, wherein the means for collecting feedback includes means for storing user ratings in a ratings database and for analyzing the data.

3. 2. The system of claim 1, wherein the character with which the user interacts is visually displayed on a terminal operated by the user.

Images