System

Abstract

A system is provided.SOLUTION: A system comprising: means for receiving a condition input by a user in a chat form using terminals; a generation AI model for generating candidate wines based on the received condition; means for transmitting a generated candidate list of wines to the terminals; means for receiving feedback provided by the user and passing the condition to the generation AI model again; means for transmitting a re-proposed candidate list of wines to the terminals; and means for generating a procedure page for purchasing wines selected by the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Choosing wine is necessary for various occasions, such as celebrations, home parties, and gifts, but for people who are not familiar with wine, it can be difficult to know how to choose and to find the perfect bottle from the many options available. This takes time and effort, and is a burden for users. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems with a system including: means for receiving conditions entered by a user in a chat format using a terminal; a generative AI model that generates candidate wines based on the received conditions; means for transmitting the generated candidate wine list to the terminal; means for receiving feedback provided by the user and again passing the conditions to the generative AI model; means for transmitting the re-proposed candidate wine list to the terminal; and means for generating a procedure page for purchasing the wine selected by the user. This allows users to efficiently and effectively find and select the wine that suits them, significantly reducing the effort and time required.

[0006] "User" refers to a person who accesses the system, inputs conditions, makes wine suggestions, and completes purchase procedures.

[0007] A "terminal" is a device used by a user, including a smartphone, tablet, or PC.

[0008] "Conditions" refers to information such as budget, preferred taste, and intended use that the user enters when selecting wine.

[0009] "Chat-style" refers to an interface in which the user and the system interact through text messages.

[0010] "Means for receiving" refers to a function of the system for acquiring conditions input by a user.

[0011] A "generative AI model" refers to an artificial intelligence algorithm that generates optimal wine candidates based on user-entered criteria.

[0012] "Shortlist" refers to a list of multiple wines suggested by the generative AI model.

[0013] "Means for sending" refers to the system's functionality for transferring the generated data (e.g., candidate list) to the user's terminal.

[0014] "Feedback" refers to additional requests or conditions that a user provides in response to the initially presented candidates.

[0015] "Means of passing conditions" refers to the system's ability to forward user feedback to the generative AI model.

[0016] The "purchase procedure page" refers to the web page or app screen where the user selects a wine and completes the purchase procedure. [Brief explanation of the drawings]

[0017] [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

[0018] 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.

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

[0020] 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).

[0021] 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.

[0022] 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.

[0023] 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.

[0024] 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."

[0025] [First embodiment]

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

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form 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.

[0032] 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.

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

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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."

[0038] This invention is a wine concierge system that utilizes a generative AI model to suggest the most suitable wine based on the user's input criteria such as budget and preferences when selecting a wine. Below, we will explain the program processing of this system in natural language and provide specific examples.

[0039] System Overview

[0040] The system is configured as follows:

[0041] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[0042] 2. The device displays a chat-style interface to the user, and the user enters their requirements (budget, preferred flavor, usage situation, etc.).

[0043] 3. The terminal sends the entered conditions to the server.

[0044] 4. The server receives the conditions and passes the data to the generative AI model.

[0045] 5. The generative AI model generates optimal wine candidates based on the criteria and sends them back to the server.

[0046] 6. The server organizes the generated candidate list and sends it to the terminal.

[0047] 7. The device displays the candidate list on the chat interface and presents it to the user.

[0048] 8. The user enters feedback on the presented options (e.g., "I'd like something a little sweeter" or "I'd be happy to increase the budget").

[0049] 9. The device sends feedback to the server, which again passes the conditions to the generative AI model.

[0050] 10. The generative AI model generates wine candidates again based on the new criteria.

[0051] 11. The server sends the new candidate list to the terminal, which again presents it to the user in the chat interface.

[0052] 12. The device will then display a page where the user can select the perfect wine and complete the purchase.

[0053] 13. Once the user completes the purchase, the server processes the payment and arranges for shipping.

[0054] Examples:

[0055] Example 1: Choosing wine for a home party

[0056] 1. The user enters the criteria

[0057] The user enters, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[0058] The terminal transmits this condition to the server as text data.

[0059] 2. The server receives the conditions and the AI ​​model generates candidates.

[0060] The server receives the conditions and passes them to the generative AI model.

[0061] The generative AI model selects three red wines from a database that meet the criteria.

[0062] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[0063] 3. Users provide feedback

[0064] User: "Do you have any sweeter wines?"

[0065] The terminal sends this feedback to the server.

[0066] 4. The server receives the feedback and the AI ​​model generates candidates again.

[0067] The server receives the feedback and again passes the new conditions to the generative AI model.

[0068] A generative AI model selects a sweet red wine.

[0069] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[0070] 5. The user makes the final decision and purchases the wine.

[0071] User: "I'll choose wine E."

[0072] The terminal transmits the selection result to the server.

[0073] The server generates a purchase page and sends it to the terminal.

[0074] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[0075] This allows users to efficiently and effectively find the wine that suits them, and through this process, users can significantly reduce the effort and time required to select a wine.

[0076] The processing flow will be explained below.

[0077] Step 1:

[0078] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[0079] Step 2:

[0080] The user inputs criteria into a chat-style interface. These criteria include "budget," "preferred flavor," and "usage occasion." The user might input something like, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[0081] Step 3:

[0082] The device sends the entered conditions as text data to the server, which then analyzes the data, converts it into an appropriate format, and prepares it for passing to the generative AI model.

[0083] Step 4:

[0084] The server receives the criteria sent by the user and passes them to the generative AI model, which searches a wine database based on the criteria and generates optimal wine candidates.

[0085] Step 5:

[0086] The generative AI model selects several wine candidates that meet the criteria, for example, three highly rated red wines within a budget of 3,000 yen.

[0087] Step 6:

[0088] The server organizes the candidate list of wines received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, etc.

[0089] Step 7:

[0090] The server sends the formatted wine candidate list to the terminal, which displays the received data on the chat screen and presents wine candidates to the user.

[0091] Step 8:

[0092] The user can input feedback on the presented wine candidates, for example, specifying additional conditions such as "I would prefer something a little sweeter" or "I'm willing to increase my budget a little."

[0093] Step 9:

[0094] The device sends the user's feedback as text data to the server, which then prepares to pass the data to the generative AI model again based on the new conditions.

[0095] Step 10:

[0096] The server receives the feedback from the user, analyzes it, and passes the new criteria to the generative AI model, which then re-searches for wine candidates based on the new criteria.

[0097] Step 11:

[0098] The generative AI model again selects wine candidates that meet the new criteria and sends them back to the server, which then organizes the list of re-suggested wines.

[0099] Step 12:

[0100] The server sends the reformatted list of wine candidates to the terminal, which then displays the received data on the chat screen and re-proposes the wine to the user.

[0101] Step 13:

[0102] The user selects the best wine from the provided candidates and sends the selection result to the server via the terminal.

[0103] Step 14:

[0104] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[0105] Step 15:

[0106] The terminal displays a checkout page to the user in a chat interface, where the user can complete the purchase.

[0107] Step 16:

[0108] Once the user completes the purchase procedure, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[0109] Step 17:

[0110] The device will notify the user of the completion of the purchase and provide information such as the estimated delivery date, allowing the user to select and purchase wine efficiently and effectively.

[0111] Example 1

[0112] 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."

[0113] Conventional wine selection systems make it difficult for users to effectively select wines that match their preferences and requirements, and the selection process takes a lot of time and effort. Furthermore, the process of reflecting user feedback and re-recommending wines is complicated and inefficient. Another problem is that purchasing procedures, payment processing, and shipping arrangements cannot be managed in a single system.

[0114] 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.

[0115] In this invention, the server includes means for receiving conditions input by a user in chat format using an information device, a generative AI model that generates candidate foods and beverages based on the received conditions, means for transmitting the generated candidate food and beverage list to the information device, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for transmitting the re-proposed candidate food and beverage list to the information device, means for generating a procedure page for purchasing the food and beverage selected by the user, means for passing the conditions received from the terminal to the generative AI model in the form of a prompt sentence, means for passing the feedback again to the generative AI model in the form of a prompt sentence when the user provides new feedback, and means for processing payment and arranging for delivery to complete the purchase procedure. This allows users to efficiently and effectively find the food and beverages that suit them and easily complete the purchase procedure.

[0116] An "information device" is a terminal that a user can access (such as a smartphone, tablet, or PC).

[0117] A "generative AI model" is an artificial intelligence algorithm that generates candidate foods and beverages based on received conditions.

[0118] A "prompt format" is a textual description of a question or instruction provided to a generative AI model.

[0119] A "candidate list" is a list of multiple foods and drinks suggested by the generative AI model.

[0120] "Feedback" refers to a reaction, including hopes, requests, corrections, etc., provided by a user to a candidate list.

[0121] "Payment processing" refers to the procedure for completing the payment required to purchase the food and drink selected by the user.

[0122] "Shipping arrangements" are logistics procedures for delivering purchased food and drink to the user.

[0123] "Chat style" refers to an interface between the user and the system that is based on the exchange of text messages.

[0124] "Conditions" are selection criteria entered by the user, such as budget, preferred flavor, and usage scenario.

[0125] The "procedure page" is a web page or app screen for the user to purchase the food and drink selected by the user.

[0126] The present invention is a concierge system that utilizes a generative AI model to suggest optimal food and drink (e.g., wine) based on user input criteria. This system is implemented primarily using information devices, a server, and a generative AI model. Specific embodiments of this system are described below.

[0127] System configuration

[0128] The system is configured as follows:

[0129] 1. Information devices (terminals):

[0130] Devices such as smartphones, tablets, and computers are used.

[0131] These terminals display a chat-style interface for users to enter requirements and provide feedback.

[0132] 2. Server:

[0133] The server implements a generative AI model.

[0134] It receives conditions and feedback entered by the user through the device and passes them on to the generative AI model.

[0135] The generated candidate list is also organized and transmitted to the terminal.

[0136] 3. Generative AI Model:

[0137] Generative AI models (e.g., OpenAI GPT-4 or Transformer-based models) generate optimal food and drink candidates based on criteria and feedback provided by the server.

[0138] The model selects suitable candidates from the database and returns this information to the server.

[0139] System operation example

[0140] As a concrete example, let us consider a scenario of selecting wine for a home party.

[0141] 1. User enters criteria:

[0142] A user enters the following into the chat interface of an information device:

[0143] "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[0144] 2. The server passes the conditions to the generative AI model:

[0145] The server receives this condition and sends it to the generative AI model in the form of a prompt sentence.

[0146] For example: "Please suggest a red wine for a home party within the budget of 3000 yen."

[0147] 3. The generative AI model generates wine candidates:

[0148] Based on the criteria, the generative AI model generates wine candidates such as:

[0149] "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[0150] 4. Users provide feedback:

[0151] The user enters feedback.

[0152] For example: "Do you have any wines that are a little sweeter?"

[0153] 5. The server passes the feedback back to the generative AI model:

[0154] The server then sends the received feedback back to the generative AI model in the form of a prompt sentence.

[0155] For example: "Please suggest a sweeter red wine."

[0156] 6. The generative AI model again generates wine candidates:

[0157] The generative AI model generates new candidates based on the new conditions, such as:

[0158] "The recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[0159] The user selects the most suitable wine from these options, and the terminal generates a purchase page. Once the purchase procedure is complete, the server processes the payment and arranges for delivery.

[0160] In this way, the present invention not only allows the user to efficiently select food and drink, but also allows the user to complete the purchase procedure in an integrated manner, thereby significantly reducing the time and effort required for selecting food and drink.

[0161] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0162] Step 1:

[0163] The user enters the conditions using a terminal.

[0164] Input: A chat interface displayed on an information device (smartphone, tablet, PC). The user types, "I'm looking for wine to use at a home party. My budget is under 3000 yen, and red wine would be good."

[0165] Output: The conditions are sent as text data from the terminal to the server.

[0166] Specific operation: The device receives user input and sends that data to the server as an API request.

[0167] Step 2:

[0168] The server receives the conditions and passes them to the generative AI model in the form of a prompt sentence.

[0169] Input: Conditional text data sent from the terminal.

[0170] Output: The conditions converted into prompt sentence format are passed to the generative AI model.

[0171] Specific operation: The server converts the condition text into a prompt sentence format and inputs it into the generative AI model. It passes it to the model as "Please suggest a red wine for a home party within a budget of 3,000 yen."

[0172] Step 3:

[0173] A generative AI model generates food and drink suggestions based on criteria.

[0174] Input: The prompt text passed by the server.

[0175] Output: A list of food and drink options (wines).

[0176] Specific operation: The generative AI model selects wines from the database that meet the conditions and generates a candidate list. For example, it selects "Wine A (2,500 yen)," "Wine B (2,800 yen)," and "Wine C (3,000 yen)."

[0177] Step 4:

[0178] The server organizes the generated candidate list and transmits it to the terminal.

[0179] Input: A candidate list generated by a generative AI model.

[0180] Output: The curated candidate list is sent to the terminal.

[0181] Specific operation: The server organizes the candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[0182] Step 5:

[0183] The terminal presents the candidate list to the user.

[0184] Input: Candidate list data sent from the server.

[0185] Output: The candidate list displayed in the chat interface.

[0186] Specific operation: The device receives the candidate list and displays it to the user in the chat interface. For example, "We recommend the following three red wines: 1. Wine A (2500 yen) 2. Wine B (2800 yen) 3. Wine C (3000 yen)."

[0187] Step 6:

[0188] The user inputs feedback on the candidate list.

[0189] Input: User-provided feedback through a chat interface. For example, "Do you have any wines that are a little sweeter?"

[0190] Output: Feedback is sent as text data from the device to the server.

[0191] Specific operation: The device receives the user's feedback and sends it to the server as an API request.

[0192] Step 7:

[0193] The server receives the feedback and passes it back to the generative AI model in the form of a prompt sentence.

[0194] Input: Feedback text data sent from the device.

[0195] Output: The feedback converted into a new prompt format is passed to the generative AI model.

[0196] Specific operation: The server converts the feedback text into a prompt sentence format and passes it to the generative AI model as "Please suggest a sweeter red wine."

[0197] Step 8:

[0198] The generative AI model then generates new food and drink suggestions based on the new criteria.

[0199] Input: The new prompt text passed by the server.

[0200] Output: A list of new candidate foods and drinks (wine).

[0201] Specific operation: The generative AI model selects sweet red wines from the database based on the new criteria and generates a candidate list. For example, it selects "Wine D (2,600 yen)" and "Wine E (2,900 yen)."

[0202] Step 9:

[0203] The server organizes a new candidate list and sends it to the terminal again.

[0204] Input: A new candidate list generated by a generative AI model.

[0205] Output: The new, cleaned candidate list is sent to the terminal.

[0206] Specific operation: The server organizes the new candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[0207] Step 10:

[0208] The terminal presents the new candidate list to the user.

[0209] Input: The new candidate list data sent from the server.

[0210] Output: The new candidate list displayed in the chat interface.

[0211] Specific operation: The device displays the new candidate list received by the device to the user in the chat interface. For example, "Recommended sweet red wines are Wine D (2600 yen) and Wine E (2900 yen)."

[0212] Step 11:

[0213] The user selects the best wine.

[0214] Input: Final candidates selected by the user through the chat interface.

[0215] Output: The selection results are sent as text data from the terminal to the server.

[0216] Specific operation: The user enters "I will choose wine E," and the device sends the selection result to the server as an API request.

[0217] Step 12:

[0218] The terminal displays a purchase page to the user.

[0219] Input: Purchase page generation instructions from the server.

[0220] Output: Purchase page displayed on the information device.

[0221] Specific operation: The server generates a link to the purchase page and sends it to the device, and the device opens the link and displays the purchase page.

[0222] Step 13:

[0223] The user completes the purchase.

[0224] Input: User's payment information on the purchase page.

[0225] Output: Confirmation that the purchase was successful.

[0226] What happens: The user enters payment information on the purchase page and completes the checkout process.

[0227] Step 14:

[0228] The server processes the payment and arranges shipping.

[0229] Input: Checkout and payment information.

[0230] Output: Confirmation that payment processing has been completed and shipping arrangements have been made.

[0231] Specific operation: The server works with the payment gateway to process the payment and uses the logistics API to arrange for shipping.

[0232] (Application example 1)

[0233] 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."

[0234] Conventional wine selection systems are limited to specific locations, such as the user's home or office. Furthermore, their reliance on smartphones or tablets limits the real-time in-store purchasing experience. Furthermore, while immediate feedback from users is required, there is a lack of efficient means to achieve this. Therefore, there is a need for a system that allows users to efficiently select the optimal wine in-store.

[0235] 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.

[0236] In this invention, the server includes means for receiving conditions input by a user in a chat format using a terminal, a generative AI model for generating candidate wines based on the received conditions, means for displaying a list of candidate wines on the smart glasses, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for displaying the list of re-suggested candidate wines on the smart glasses, and means for generating a procedure page for purchasing the wine selected by the user. This allows the user to select the optimal wine in real time in the store and quickly receive re-suggested wines based on the feedback.

[0237] "User" means an individual or group that uses a service or product.

[0238] A "terminal" is an electronic device used by a user to input conditions, such as a smartphone, tablet, PC, or smart glasses.

[0239] "Chat style" refers to an interface style in which users exchange text and voice in a natural conversational flow.

[0240] "Conditions" refer to specific requests or requirements entered by the user (e.g., budget, preferred flavor, usage scenario, etc.).

[0241] "Receiving" refers to the act of receiving input data or information.

[0242] A "generative AI model" is an artificial intelligence model that is trained to generate optimal candidates based on a user's criteria.

[0243] A "candidate list" is a list of multiple options selected by a generative AI model.

[0244] "Feedback" refers to opinions and additional requests provided by users.

[0245] "Re-proposal" refers to the act of presenting a newly selected candidate based on the feedback received.

[0246] The "procedure page" is a web page where the user inputs and confirms the information required to purchase the selected wine.

[0247] "Smart glasses" are electronic devices in the form of glasses that have a built-in display and camera and have the ability to display information superimposed on the user's field of vision.

[0248] "Display" refers to the act of visually presenting information on the screen of a device or smart glasses.

[0249] "Purchase" refers to the act of obtaining goods or services in exchange for money.

[0250] This invention provides a wine concierge system for efficiently selecting wines that meet the user's needs. Specifically, the user inputs criteria using smart glasses, and a generative AI model generates an optimal candidate list. Based on the feedback, a new candidate list of wines is then displayed, allowing the user to select the optimal wine and proceed with the purchase.

[0251] System Configuration

[0252] The system consists of the following main components:

[0253] 1. User Device:

[0254] A device that allows users to input conditions, including smart glasses. Smart glasses display a chat-style interface and allow voice or touch input. Google Glass is a typical example.

[0255] 2. Server:

[0256] It receives conditions sent from the user's device and passes them to the generative AI model. The server runs in a cloud environment (e.g., an EC2 instance on AWS). Here, it is built as a web application using Flask.

[0257] 3. Generative AI Model:

[0258] It is a generative AI model trained using PyTorch to generate optimal wine candidates based on user input criteria. Specifically, it uses the GPT-3 model.

[0259] Program processing

[0260] The server receives conditional input from the user using a Flask application.

[0261] The received criteria are passed to the GPT-3 model to generate an optimal wine candidate list.

[0262] The server sends the generated wine candidate list to the smart glasses, which then overlays it on the field of view.

[0263] When feedback is received, the conditions are passed to the generative AI model again, and a regenerated candidate list is displayed on the smart glasses.

[0264] Finally, a procedure page for purchasing the wine selected by the user is generated and displayed on the smart glasses.

[0265] Specific examples

[0266] When a user wants to select wine for a house party, the following steps occur:

[0267] 1. Enter the conditions:

[0268] The user voice-inputs into the chat interface of the smart glasses, "I'm looking for wine to use at a home party. My budget is under 4,000 yen, and I'd prefer a white wine."

[0269] 2. Generate wine candidates:

[0270] The server passes the received criteria to the GPT-3 model, which generates an optimal wine candidate list based on the criteria.

[0271] 3. Displaying the results:

[0272] A list of wine candidates is sent from the server to the smart glasses and displayed over the user's field of view. For example, it might look like this: "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[0273] 4. Providing Feedback:

[0274] When the user voices feedback such as "Do you have any sweeter wines?", the server receives this and passes the new criteria to the generative AI model again to generate a revised list of candidates.

[0275] 5. View resubmissions:

[0276] The regenerated candidate list is displayed on the smart glasses. For example, it may look like this: "Candidate 1: Wine C (3,600 yen, sweet)."

[0277] 6. Purchase Procedure:

[0278] The user selects the most suitable wine and inputs the selection result by saying, "I will choose wine C." Based on this, the server generates a checkout page and displays it on the smart glasses.

[0279] Prompt Sentence Examples

[0280] As a concrete example, the prompt sentence "Home party wine budget 4000 yen white wine" is passed to the generative AI model. Based on this prompt sentence, the generative AI model will suggest the most suitable wine candidates.

[0281] The system allows users to use smart glasses to select the perfect wine in real time and quickly receive additional suggestions based on feedback.

[0282] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0283] Step 1:

[0284] The user inputs the requirements by voice into the chat interface of the smart glasses. In this step, the voice is converted into text using a speech recognition API and stored in a database. The specific input is "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable." The output is the textual requirements: "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable."

[0285] Step 2:

[0286] The device sends the textual conditions to the cloud server, which then receives the conditions and prepares them for passing to the generative AI model. The input is the textual conditions, and the output is an HTTP request containing the conditions.

[0287] Step 3:

[0288] The server passes the received conditions to a generative AI model to generate a list of wine candidates. Specifically, it uses a GPT-3 model to generate the best wine candidates based on the conditions. The input is text conditions, and the conditions are converted into prompt sentences as data processing and input into the generative AI model. The output is a list of wine candidates. For example, "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[0289] Step 4:

[0290] The server sends the generated wine candidate list to the smart glasses and displays it over the user's field of view. In this step, data is sent via an HTTP response and the information is displayed on the smart glasses' display. The input is the wine candidate list, and the output is the information overlaid on the smart glasses' display.

[0291] Step 5:

[0292] The user speaks their feedback to the displayed list of candidates, and the feedback is converted into text. For example, they can say, "Do you have any sweeter wines?" and the speech recognition API is used to convert it into text. The input is spoken feedback, and the output is text feedback.

[0293] Step 6:

[0294] The device sends textual feedback to the server, which receives and analyzes it. The conditions are passed to the generative AI model again to generate a new candidate list. The input is the textual feedback, which is converted into a new prompt sentence as data processing and input to the generative AI model. The output is a regenerated list of wine candidates. For example, "Candidate 1: Wine C (3,600 yen, sweet)."

[0295] Step 7:

[0296] The server sends the regenerated wine candidate list to the smart glasses and displays it over the user's field of view. The input is the regenerated wine candidate list, and the output is the information overlaid on the smart glasses' display.

[0297] Step 8:

[0298] The user selects the most suitable wine and inputs the selection by voice or touch. For example, "I will choose wine C." The input is the user's selection, and the output is a text version of the selection.

[0299] Step 9:

[0300] The terminal sends the selection result to the server, and the server generates a checkout page. In this step, a web page containing the information necessary for the checkout process is generated based on the selection result and sent to the smart glasses. The input is the selection result, and the output is the checkout page.

[0301] Step 10:

[0302] The user enters the necessary information on the checkout page displayed on the smart glasses to complete the purchase. Specifically, they enter credit card information and shipping address information. The input is the user's purchase procedure information, and the output is a confirmation message that the purchase has been completed.

[0303] Through these processing steps, users can select the best wine in real time and quickly receive additional suggestions based on feedback.

[0304] 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.

[0305] This invention combines a wine concierge system that utilizes a generative AI model that suggests the most suitable wine based on user input criteria such as budget and preferences when selecting a wine, with an emotion engine that recognizes the user's emotions. Below, the program processing of this system is explained in natural language, with specific examples.

[0306] System Overview

[0307] The system is configured as follows:

[0308] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[0309] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred flavor, usage scenario, etc.).

[0310] 3. The terminal sends the entered conditions to the server as text data.

[0311] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[0312] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[0313] 6. The generative AI model searches a database of wines based on criteria and sentiment to generate the best wine candidates.

[0314] 7. The server organizes the generated candidate list and sends it to the terminal.

[0315] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[0316] 9. The user enters feedback on the presented wine candidates (e.g., "I'd like something sweeter" or "I'm willing to increase my budget")

[0317] 10. The device sends the feedback to the server.

[0318] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[0319] 12. The generative AI model again generates wine suggestions based on criteria and sentiment.

[0320] 13. The server sends the re-proposal candidate list to the terminal.

[0321] 14. The device again presents the chat interface to the user.

[0322] 15. The device will then display a page where the user can select the perfect wine and complete the purchase.

[0323] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[0324] Examples:

[0325] Example 1: Choosing wine for a home party

[0326] 1. The user enters the criteria

[0327] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[0328] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[0329] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[0330] The server receives the conditions and emotion data and passes them to the generative AI model.

[0331] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[0332] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[0333] 3. Users provide feedback

[0334] User: "Do you have any sweeter wines?"

[0335] The device sends this feedback and the user's emotional data to the server.

[0336] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[0337] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[0338] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[0339] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[0340] 5. The user makes the final decision and purchases the wine.

[0341] User: "I'll choose wine E."

[0342] The terminal transmits the selection result to the server.

[0343] The server generates a purchase page and sends it to the terminal.

[0344] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[0345] This allows users to find the perfect wine that reflects their emotions. Through this process, users can significantly reduce the effort and time required to select a wine while also improving their emotional satisfaction.

[0346] The processing flow will be explained below.

[0347] Step 1:

[0348] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[0349] Step 2:

[0350] The user enters the conditions into a chat-style interface. For example, the user might enter, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device then sends this condition to the server as text data.

[0351] Step 3:

[0352] The server receives the conditions. At the same time, the emotion engine analyzes the emotional data from the user's input and generates emotional information such as "joy" or "expectation." This data is then passed to the generative AI model by the server.

[0353] Step 4:

[0354] A generative AI model searches a database of wines based on criteria (budget, red wine) and emotional data (e.g., joy, anticipation). The AI ​​model generates optimal wine candidates and sends them back to the server.

[0355] Step 5:

[0356] The server organizes the wine recommendation list received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, and emotional reasons for the recommendation.

[0357] Step 6:

[0358] The terminal displays the formatted wine candidate list on the chat interface and presents it to the user, who can then confirm the presented candidate list.

[0359] Step 7:

[0360] The user inputs feedback for the presented wine candidates. For example, the user might input, "Do you have any wines that are a little sweeter?" The device then sends this feedback to the server.

[0361] Step 8:

[0362] The server receives the user feedback and analyzes the emotion data again through the emotion engine, which prepares to pass the updated emotion information to the generative AI model.

[0363] Step 9:

[0364] The server passes the feedback and updated emotional data to the generative AI model, which then re-searches for wine candidates based on the new criteria (sweet red wine) and the updated emotional data.

[0365] Step 10:

[0366] The generative AI model generates new wine candidates based on the new criteria and emotional data and sends them back to the server, which receives them and organizes the list of candidates for re-suggestion.

[0367] Step 11:

[0368] The terminal displays the re-suggested wine candidate list in the chat interface and presents it to the user. The re-suggested candidate list includes information based on the new conditions and emotional information.

[0369] Step 12:

[0370] The user selects the best wine. For example, the user inputs, "I will choose wine E." The device sends this selection result to the server.

[0371] Step 13:

[0372] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[0373] Step 14:

[0374] The terminal displays a purchase procedure page to the user using a chat interface, where the user enters the necessary information and completes the purchase procedure.

[0375] Step 15:

[0376] Once the user completes the purchase, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[0377] Step 16:

[0378] The device will display a notification to the user that the purchase is complete and provide information such as the estimated delivery date, allowing the user to select wine efficiently and emotionally in a satisfying way.

[0379] Example 2

[0380] 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."

[0381] Conventional wine selection systems only consider the criteria entered by the user to suggest products, which tends to result in suggestions that ignore the user's feelings. This reduces user satisfaction and makes it difficult to select the appropriate product. In addition, the process of re-suggesting products based on feedback often does not function efficiently, resulting in a suboptimal user experience.

[0382] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing emotions from user input data, means for transferring the analyzed emotion data to the information processing model, and means for analyzing feedback provided by the user and transferring conditions based on the feedback back to the information processing model. This makes it possible to propose products that reflect the user's emotions, improving user satisfaction and realizing an efficient feedback re-proposal process.

[0383] A "user" is a user who uses a terminal to access the service and input conditions and feedback.

[0384] "Terminal" refers to an electronic device used by a user, such as a smartphone, tablet, or PC, and is a device for inputting conditions and feedback through a chat-style interface.

[0385] The "server" is a central system that processes data received from users, performs sentiment analysis, product suggestions and re-suggestions, and sends the results to the terminal.

[0386] The "conditions" refer to the user's wishes and requirements regarding the product he or she wishes to purchase, and specifically include budget, preferred type, purpose, and the like.

[0387] An "information processing model" is an algorithm or system that generates optimal product candidates based on user-entered conditions and sentiment analysis data.

[0388] A "generative AI model" is a type of information processing model, a system that uses AI technology to generate product candidates based on user conditions and emotional data.

[0389] An "emotion engine" is a system that analyzes emotions from user input data and provides the results to a generative AI model.

[0390] "Feedback" refers to additional opinions or requests provided by the user regarding the proposed product candidates, and serves as reference information when the system makes another suggestion.

[0391] A "product candidate list" is a list of candidate products that are judged to meet the conditions from among the group of products proposed by the generative AI model.

[0392] "Re-proposal" refers to the process in which the generative AI model regenerates and proposes a list of product candidates based on user feedback.

[0393] This invention combines an emotion engine that recognizes the user's emotions with a product concierge system that utilizes an information processing model that suggests optimal products based on user input criteria such as budget and preferences when selecting a product. The program processing of this system is explained below, along with specific examples.

[0394] System Overview

[0395] The system is configured as follows:

[0396] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[0397] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred features, usage scenario, etc.).

[0398] 3. The terminal sends the entered conditions to the server as text data.

[0399] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[0400] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[0401] 6. The generative AI model searches the product database based on the conditions and sentiment to generate the best product candidates.

[0402] 7. The server organizes the generated candidate list and sends it to the terminal.

[0403] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[0404] 9. The user enters feedback on the presented product candidates (e.g., "I'd like something sweeter" or "I'd be okay with a higher budget").

[0405] 10. The device sends the feedback to the server.

[0406] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[0407] 12. The generative AI model again generates product candidates based on the conditions and sentiment.

[0408] 13. The server sends the re-proposal candidate list to the terminal.

[0409] 14. The device again presents the chat interface to the user.

[0410] 15. A page will be displayed on the device that allows the user to select the most suitable product and complete the purchase process.

[0411] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[0412] Hardware and software used

[0413] Users use devices such as smartphones, tablets, and PCs.

[0414] The servers are high-performance cloud servers equipped with data processing and storage capabilities.

[0415] The emotion engine uses natural language processing technology to analyze emotions from user input data.

[0416] Generative AI models use machine learning algorithms to suggest the best products based on criteria and sentiment.

[0417] Specific examples

[0418] Example 1: Choosing wine for a home party

[0419] 1. The user enters the criteria

[0420] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[0421] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[0422] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[0423] The server receives the conditions and emotion data and passes them to the generative AI model.

[0424] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[0425] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)" is sent to the terminal.

[0426] 3. Users provide feedback

[0427] User: "Do you have any sweeter wines?"

[0428] The device sends this feedback and the user's emotional data to the server.

[0429] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[0430] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[0431] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[0432] Server: Sends the message "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)" to the terminal.

[0433] 5. The user makes the final decision and purchases the wine.

[0434] User: "I'll choose wine E."

[0435] The terminal transmits the selection result to the server.

[0436] The server generates a purchase page and sends it to the terminal.

[0437] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[0438] This entire process allows users to find the perfect product that reflects their emotions and complete the purchase process smoothly.

[0439] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0440] Step 1:

[0441] The user enters the conditions

[0442] Input: Budget, preferred features, usage scenario, etc.

[0443] Output: Text data containing the condition

[0444] Specific operation: The user opens a chat interface on the device and enters criteria in text format, such as "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device captures this input as text data and immediately sends it to the server.

[0445] Step 2:

[0446] The device sends the user-entered data to the server

[0447] Input: Text data of the condition entered by the user

[0448] Output: Text data sent to the server

[0449] Specific operation: The terminal sends the user's input data to the server via secure communication. The input data is converted into the required format and then passed to the server.

[0450] Step 3:

[0451] The server analyzes the condition and emotion

[0452] Input: Condition data received from the user

[0453] Output: Condition data and emotion data

[0454] Specific operation: The server passes the received text data to the emotion engine, which analyzes the emotions (joy, anticipation, etc.) from the user's input. Condition data and emotion data are generated as the analysis results.

[0455] Step 4:

[0456] The server passes the analysis results to the generative AI model

[0457] Input: Condition data and emotion data

[0458] Output: The data fed into the generative AI model

[0459] Specific operation: The server integrates the emotion data obtained by the emotion engine with the user's condition data and passes it to the generative AI model.

[0460] Step 5:

[0461] Generative AI model generates product candidates

[0462] Input: Condition data and emotion data

[0463] Output: Product candidate list

[0464] Specific operation: The generative AI model searches the database based on the conditions and emotions. For example, if the conditions are "red wine, budget under 3,000 yen," it retrieves appropriate products (e.g., Wine A, Wine B, Wine C) from the database and generates a product candidate list.

[0465] Step 6:

[0466] The server sends the generated candidate list to the device.

[0467] Input: Product candidate list

[0468] Output: Product candidate list sent to the terminal

[0469] Specific operation: The server organizes the generated candidate list, converts it into an easy-to-understand format, and then sends the candidate list to the terminal.

[0470] Step 7:

[0471] The device presents a candidate list to the user.

[0472] Input: Product candidate list sent from the server

[0473] Output: Product candidate list displayed in chat interface

[0474] Specific operation: The device presents the user with a list of options via a chat-style interface, such as "We recommend the following three products: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)."

[0475] Step 8:

[0476] User enters feedback

[0477] Input: Feedback on product candidate list

[0478] Output: Feedback text data

[0479] Specific operation: The user inputs feedback such as "Do you have any wines that are a little sweeter?", and the device captures this feedback as text data and sends it to the server.

[0480] Step 9:

[0481] The device sends feedback to the server

[0482] Input: User feedback text data

[0483] Output: Feedback data sent to the server

[0484] How it works: The device sends feedback to the server over secure communication, and the data is converted into the required format and then passed to the server.

[0485] Step 10:

[0486] The server analyzes the emotions and feedback again.

[0487] Input: Feedback data

[0488] Output: Updated condition data and emotion data

[0489] Specific operation: The server passes the feedback to the emotion engine, which analyzes the emotion again and updates the condition data based on the feedback. For example, if the user's feedback is "sweeter," the condition data is updated to "sweet red wine."

[0490] Step 11:

[0491] The generative AI model generates new product candidates

[0492] Input: Updated condition data and emotion data

[0493] Output: Regenerated product candidate list

[0494] Specific operation: The generative AI model re-searches the database based on new criteria (e.g., sweet red wine) and emotion data to generate new product candidates. If the new list is "Wine D (2,600 yen), Wine E (2,900 yen)," etc., it creates that list.

[0495] Step 12:

[0496] The server sends a list of re-proposals to the device.

[0497] Input: Regenerated product candidate list

[0498] Output: Regeneration candidate list sent to the terminal

[0499] Specific operation: The server organizes the regenerated candidate list, converts it into an easy-to-understand format, and sends it to the terminal.

[0500] Step 13:

[0501] The device presents the candidate list to the user again.

[0502] Input: Regeneration candidate list sent from the server

[0503] Output: Regeneration candidate list displayed in the chat interface

[0504] Specific operation: The device presents the user with a list of regeneration candidates via a chat-style interface, such as "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[0505] Step 14:

[0506] The user decides on the best product and purchases it.

[0507] Input: Final selected product

[0508] Output: Purchase procedure page generated and payment and shipping arrangements completed

[0509] Specific operation: The user enters "I'll choose wine E," and the device sends the selection result to the server. The server generates a purchase procedure page and sends it to the device. The user completes the purchase procedure, and payment and shipping arrangements are completed.

[0510] (Application example 2)

[0511] 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."

[0512] Conventional wine recommendation systems have been unable to select wines that reflect the user's emotions, making it difficult to recommend the best wine for the user. Furthermore, when selecting wine in a brick-and-mortar store, there is a lack of a way for users to efficiently input their budget and preferences and obtain the best wine in real time. The present invention aims to solve these problems and provide a system that can make better wine recommendations that reflect the user's emotions.

[0513] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving conditions entered by a user in a chat format using a terminal; a generative AI model that generates candidate wines based on the received conditions and the user's emotional data; means for transmitting the generated candidate wine list to the terminal; means for receiving feedback provided by the user and again passing the conditions to the generative AI model; means for transmitting the re-proposed candidate wine list to the terminal; means for generating a procedure page for purchasing the wine selected by the user; means for recognizing the user's emotions; means for transferring the emotional data to the generative AI model; and means for using a display device to present the wine selection results to the user. This enables optimal wine recommendations that reflect the user's emotions, enabling efficient wine selection even in physical stores.

[0514] "Means for receiving conditions entered by a user in chat format using a terminal" refers to a mechanism by which a user enters wine selection conditions in chat format using a device such as a smartphone or tablet, and receives that input.

[0515] A "generative AI model that generates candidate wines based on received conditions and user emotional data" is an artificial intelligence model that generates optimal wine candidates based on conditions and emotional data entered by the user.

[0516] The "means for transmitting the generated wine candidate list to the terminal" is a communication function for transmitting the wine candidate list generated by the AI ​​model to the user's device.

[0517] "Means for receiving feedback provided by the user and passing conditions back to the generative AI model" refers to a mechanism for receiving feedback from the user and providing conditions back to the generative AI model based on that feedback.

[0518] The "means for sending a re-proposed wine candidate list to the terminal" is a function for sending a re-generated wine candidate list generated by the AI ​​model based on the updated conditions to the user's device.

[0519] The "means for generating a procedure page for purchasing the wine selected by the user" is a system for generating a web page for purchasing the wine finally selected by the user.

[0520] "Means for recognizing user emotions" refers to technology that analyzes the user's voice, facial expressions, etc. to identify their emotions.

[0521] "Means for transferring emotion data to the generative AI model" refers to a function for transmitting emotion data obtained by emotion recognition to the generative AI model.

[0522] The "means of using a display device to present wine selection results to a user" refers to a technology that uses smart glasses or other display devices to display wine selection results to a user.

[0523] A system for implementing the present invention uses the following hardware and software.

[0524] Hardware:

[0525] Smart devices (smartphones, tablets)

[0526] Smart glasses (e.g., Google Glass)

[0527] server

[0528] software:

[0529] Python

[0530] Flask (web server framework)

[0531] OpenAI API (generative AI model)

[0532] Emotion analysis software (e.g., voice and facial expression recognition libraries)

[0533] System Overview

[0534] The system is configured as follows:

[0535] 1. User input:

[0536] Using a smart device or smart glasses, users can input wine selection criteria (e.g., budget, preferred flavor, intended use, etc.) in a chat format. The device then sends these criteria as data to the server.

[0537] 2. Collecting Emotional Data:

[0538] The smart glasses analyze the user's voice and facial expressions in real time to collect emotional data, which is then sent to the server along with the condition data.

[0539] 3. Sentiment Analysis and Conditional Processing:

[0540] The server analyzes the received conditions and emotion data and uses emotion analysis software to understand the user's emotions.

[0541] 4. Wine selection using generative AI models:

[0542] The server inputs a prompt sentence into the generative AI model based on the emotion data and condition data, and generates wine candidates. For example, the prompt sentence might look like this:

[0543] User sentiment: positive

[0544] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[0545] Choose the three best wines.

[0546] 5. User suggestions for wines:

[0547] The server generates a list of wine candidates, which is then sent to the smart device or smart glasses and displayed to the user. For example, it might show "Wine A (2,500 yen), Wine B (2,800 yen), Wine C (3,000 yen)."

[0548] 6. Feedback and Resubmissions:

[0549] The user inputs feedback on the presented wine candidates (e.g., "A slightly sweeter wine"), and the device sends this feedback and the user's emotional data to the server. The server then performs emotional analysis again, passes new conditions to the generative AI model, and generates a new list of wine candidates, which it then presents to the user.

[0550] 7. Final Decision and Purchase Procedure:

[0551] Once the user selects the wine that best suits them, the device sends the selection results to the server, which then generates a purchase procedure page and displays it to the user. The user completes the purchase procedure on that page, and payment and shipping arrangements are completed.

[0552] Examples:

[0553] Wine selection for a home party:

[0554] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[0555] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[0556] User: "Do you have any sweeter wines?"

[0557] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[0558] User: "I'll choose wine E."

[0559] This system allows users to find the perfect wine that reflects their own feelings, and also streamlines the selection process in physical stores.

[0560] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0561] Step 1:

[0562] The user uses a smart device or smart glasses to input wine selection criteria (e.g., budget, preferred taste, intended use, etc.) in chat format. This generates input data (criteria) from the user.

[0563] Step 2:

[0564] The terminal receives the data entered by the user and sends it to the server. At this time, the data is sent in text format. An example of data sent from the terminal to the server is "Budget: under 3000 yen, red wine for a home party."

[0565] Step 3:

[0566] The smart glasses analyze the user's voice and facial expressions in real time to generate emotional data. The emotional data is analyzed as expressions such as "happiness" or "expectation" using voice and facial recognition software. The generated emotional data is then sent back to the server.

[0567] Step 4:

[0568] The server analyzes the received condition data and emotion data and generates an analysis result using emotion analysis software. Specifically, a voice analysis engine and a facial expression analysis engine are used. Based on this, the user's emotional state is determined to be "positive."

[0569] Step 5:

[0570] The server combines the condition data and emotion data to generate a prompt sentence, which is then input to the generative AI model. An example of this prompt sentence is as follows:

[0571] User sentiment: positive

[0572] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[0573] Choose the three best wines.

[0574] The generative AI model generates a list of optimal wines based on this prompt.

[0575] Step 6:

[0576] The server receives the wine candidate list output from the generative AI model and sends it to the terminal. An example of a wine candidate list that is sent is "Wine A (2500 yen), Wine B (2800 yen), Wine C (3000 yen)."

[0577] Step 7:

[0578] The device receives the wine candidate list and presents it to the user in a chat format. The user then checks the candidate list displayed on the display of the smart glasses or smart device.

[0579] Step 8:

[0580] The user inputs feedback on the wine candidate list, for example, "Do you have any sweeter wines?" This feedback is sent to the server via the terminal.

[0581] Step 9:

[0582] The server analyzes the feedback data and the new emotion data, and again inputs new condition prompts into the generative AI model, which generates a revised wine candidate list. For example, the following prompts are used:

[0583] User sentiment: positive

[0584] Wine selection criteria: A sweet red wine for a home party, within a budget of 3,000 yen

[0585] Choose the three best wines.

[0586] Step 10:

[0587] The server receives the re-proposed wine candidate list and sends it to the terminal. An example of the re-proposed list sent is "Wine D (2600 yen), Wine E (2900 yen)".

[0588] Step 11:

[0589] The terminal again presents the user with a list of wine candidates, and the user selects the most suitable wine. The user then makes a final selection, such as "I will choose wine E."

[0590] Step 12:

[0591] The server generates a procedure page for purchasing the wine selected by the user and sends it to the terminal. The user completes the purchase procedure on the page, and once payment is complete, shipping arrangements are made.

[0592] 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.

[0593] 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.

[0594] 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.

[0595] [Second embodiment]

[0596] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0597] 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.

[0598] 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).

[0599] 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.

[0600] 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.

[0601] 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).

[0602] 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.

[0603] 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.

[0604] 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.

[0605] 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.

[0606] 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.

[0607] 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."

[0608] This invention is a wine concierge system that utilizes a generative AI model to suggest the most suitable wine based on the user's input criteria such as budget and preferences when selecting a wine. Below, we will explain the program processing of this system in natural language and provide specific examples.

[0609] System Overview

[0610] The system is configured as follows:

[0611] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[0612] 2. The device displays a chat-style interface to the user, and the user enters their requirements (budget, preferred flavor, usage situation, etc.).

[0613] 3. The terminal sends the entered conditions to the server.

[0614] 4. The server receives the conditions and passes the data to the generative AI model.

[0615] 5. The generative AI model generates optimal wine candidates based on the criteria and sends them back to the server.

[0616] 6. The server organizes the generated candidate list and sends it to the terminal.

[0617] 7. The device displays the candidate list on the chat interface and presents it to the user.

[0618] 8. The user enters feedback on the presented options (e.g., "I'd like something a little sweeter" or "I'd be happy to increase the budget").

[0619] 9. The device sends feedback to the server, which again passes the conditions to the generative AI model.

[0620] 10. The generative AI model generates wine candidates again based on the new criteria.

[0621] 11. The server sends the new candidate list to the terminal, which again presents it to the user in the chat interface.

[0622] 12. The device will then display a page where the user can select the perfect wine and complete the purchase.

[0623] 13. Once the user completes the purchase, the server processes the payment and arranges for shipping.

[0624] Examples:

[0625] Example 1: Choosing wine for a home party

[0626] 1. The user enters the criteria

[0627] The user enters, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[0628] The terminal transmits this condition to the server as text data.

[0629] 2. The server receives the conditions and the AI ​​model generates candidates.

[0630] The server receives the conditions and passes them to the generative AI model.

[0631] The generative AI model selects three red wines from a database that meet the criteria.

[0632] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[0633] 3. Users provide feedback

[0634] User: "Do you have any sweeter wines?"

[0635] The terminal sends this feedback to the server.

[0636] 4. The server receives the feedback and the AI ​​model generates candidates again.

[0637] The server receives the feedback and again passes the new conditions to the generative AI model.

[0638] A generative AI model selects a sweet red wine.

[0639] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[0640] 5. The user makes the final decision and purchases the wine.

[0641] User: "I'll choose wine E."

[0642] The terminal transmits the selection result to the server.

[0643] The server generates a purchase page and sends it to the terminal.

[0644] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[0645] This allows users to efficiently and effectively find the wine that suits them, and through this process, users can significantly reduce the effort and time required to select a wine.

[0646] The processing flow will be explained below.

[0647] Step 1:

[0648] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[0649] Step 2:

[0650] The user inputs criteria into a chat-style interface. These criteria include "budget," "preferred flavor," and "usage occasion." The user might input something like, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[0651] Step 3:

[0652] The device sends the entered conditions as text data to the server, which then analyzes the data, converts it into an appropriate format, and prepares it for passing to the generative AI model.

[0653] Step 4:

[0654] The server receives the criteria sent by the user and passes them to the generative AI model, which searches a wine database based on the criteria and generates optimal wine candidates.

[0655] Step 5:

[0656] The generative AI model selects several wine candidates that meet the criteria, for example, three highly rated red wines within a budget of 3,000 yen.

[0657] Step 6:

[0658] The server organizes the candidate list of wines received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, etc.

[0659] Step 7:

[0660] The server sends the formatted wine candidate list to the terminal, which displays the received data on the chat screen and presents wine candidates to the user.

[0661] Step 8:

[0662] The user can input feedback on the presented wine candidates, for example, specifying additional conditions such as "I would prefer something a little sweeter" or "I'm willing to increase my budget a little."

[0663] Step 9:

[0664] The device sends the user's feedback as text data to the server, which then prepares to pass the data to the generative AI model again based on the new conditions.

[0665] Step 10:

[0666] The server receives the feedback from the user, analyzes it, and passes the new criteria to the generative AI model, which then re-searches for wine candidates based on the new criteria.

[0667] Step 11:

[0668] The generative AI model again selects wine candidates that meet the new criteria and sends them back to the server, which then organizes the list of re-suggested wines.

[0669] Step 12:

[0670] The server sends the reformatted list of wine candidates to the terminal, which then displays the received data on the chat screen and re-proposes the wine to the user.

[0671] Step 13:

[0672] The user selects the best wine from the provided candidates and sends the selection result to the server via the terminal.

[0673] Step 14:

[0674] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[0675] Step 15:

[0676] The terminal displays a checkout page to the user in a chat interface, where the user can complete the purchase.

[0677] Step 16:

[0678] Once the user completes the purchase procedure, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[0679] Step 17:

[0680] The device will notify the user of the completion of the purchase and provide information such as the estimated delivery date, allowing the user to select and purchase wine efficiently and effectively.

[0681] Example 1

[0682] 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."

[0683] Conventional wine selection systems make it difficult for users to effectively select wines that match their preferences and requirements, and the selection process takes a lot of time and effort. Furthermore, the process of reflecting user feedback and re-recommending wines is complicated and inefficient. Another problem is that purchasing procedures, payment processing, and shipping arrangements cannot be managed in a single system.

[0684] 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.

[0685] In this invention, the server includes means for receiving conditions input by a user in chat format using an information device, a generative AI model that generates candidate foods and beverages based on the received conditions, means for transmitting the generated candidate food and beverage list to the information device, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for transmitting the re-proposed candidate food and beverage list to the information device, means for generating a procedure page for purchasing the food and beverage selected by the user, means for passing the conditions received from the terminal to the generative AI model in the form of a prompt sentence, means for passing the feedback again to the generative AI model in the form of a prompt sentence when the user provides new feedback, and means for processing payment and arranging for delivery to complete the purchase procedure. This allows users to efficiently and effectively find the food and beverages that suit them and easily complete the purchase procedure.

[0686] An "information device" is a terminal that a user can access (such as a smartphone, tablet, or PC).

[0687] A "generative AI model" is an artificial intelligence algorithm that generates candidate foods and beverages based on received conditions.

[0688] A "prompt format" is a textual description of a question or instruction provided to a generative AI model.

[0689] A "candidate list" is a list of multiple foods and drinks suggested by the generative AI model.

[0690] "Feedback" refers to a reaction, including hopes, requests, corrections, etc., provided by a user to a candidate list.

[0691] "Payment processing" refers to the procedure for completing the payment required to purchase the food and drink selected by the user.

[0692] "Shipping arrangements" are logistics procedures for delivering purchased food and drink to the user.

[0693] "Chat style" refers to an interface between the user and the system that is based on the exchange of text messages.

[0694] "Conditions" are selection criteria entered by the user, such as budget, preferred flavor, and usage scenario.

[0695] The "procedure page" is a web page or app screen for the user to purchase the food and drink selected by the user.

[0696] The present invention is a concierge system that utilizes a generative AI model to suggest optimal food and drink (e.g., wine) based on user input criteria. This system is implemented primarily using information devices, a server, and a generative AI model. Specific embodiments of this system are described below.

[0697] System configuration

[0698] The system is configured as follows:

[0699] 1. Information devices (terminals):

[0700] Devices such as smartphones, tablets, and computers are used.

[0701] These terminals display a chat-style interface for users to enter requirements and provide feedback.

[0702] 2. Server:

[0703] The server implements a generative AI model.

[0704] It receives conditions and feedback entered by the user through the device and passes them on to the generative AI model.

[0705] The generated candidate list is also organized and transmitted to the terminal.

[0706] 3. Generative AI Model:

[0707] Generative AI models (e.g., OpenAI GPT-4 or Transformer-based models) generate optimal food and drink candidates based on criteria and feedback provided by the server.

[0708] The model selects suitable candidates from the database and returns this information to the server.

[0709] System operation example

[0710] As a concrete example, let us consider a scenario of selecting wine for a home party.

[0711] 1. User enters criteria:

[0712] A user enters the following into the chat interface of an information device:

[0713] "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[0714] 2. The server passes the conditions to the generative AI model:

[0715] The server receives this condition and sends it to the generative AI model in the form of a prompt sentence.

[0716] For example: "Please suggest a red wine for a home party within the budget of 3000 yen."

[0717] 3. The generative AI model generates wine candidates:

[0718] Based on the criteria, the generative AI model generates wine candidates such as:

[0719] "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[0720] 4. Users provide feedback:

[0721] The user enters feedback.

[0722] For example: "Do you have any wines that are a little sweeter?"

[0723] 5. The server passes the feedback back to the generative AI model:

[0724] The server then sends the received feedback back to the generative AI model in the form of a prompt sentence.

[0725] For example: "Please suggest a sweeter red wine."

[0726] 6. The generative AI model again generates wine candidates:

[0727] The generative AI model generates new candidates based on the new conditions, such as:

[0728] "The recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[0729] The user selects the most suitable wine from these options, and the terminal generates a purchase page. Once the purchase procedure is complete, the server processes the payment and arranges for delivery.

[0730] In this way, the present invention not only allows the user to efficiently select food and drink, but also allows the user to complete the purchase procedure in an integrated manner, thereby significantly reducing the time and effort required for selecting food and drink.

[0731] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0732] Step 1:

[0733] The user enters the conditions using a terminal.

[0734] Input: A chat interface displayed on an information device (smartphone, tablet, PC). The user types, "I'm looking for wine to use at a home party. My budget is under 3000 yen, and red wine would be good."

[0735] Output: The conditions are sent as text data from the terminal to the server.

[0736] Specific operation: The device receives user input and sends that data to the server as an API request.

[0737] Step 2:

[0738] The server receives the conditions and passes them to the generative AI model in the form of a prompt sentence.

[0739] Input: Conditional text data sent from the terminal.

[0740] Output: The conditions converted into prompt sentence format are passed to the generative AI model.

[0741] Specific operation: The server converts the condition text into a prompt sentence format and inputs it into the generative AI model. It passes it to the model as "Please suggest a red wine for a home party within a budget of 3,000 yen."

[0742] Step 3:

[0743] A generative AI model generates food and drink suggestions based on criteria.

[0744] Input: The prompt text passed by the server.

[0745] Output: A list of food and drink options (wines).

[0746] Specific operation: The generative AI model selects wines from the database that meet the conditions and generates a candidate list. For example, it selects "Wine A (2,500 yen)," "Wine B (2,800 yen)," and "Wine C (3,000 yen)."

[0747] Step 4:

[0748] The server organizes the generated candidate list and transmits it to the terminal.

[0749] Input: A candidate list generated by a generative AI model.

[0750] Output: The curated candidate list is sent to the terminal.

[0751] Specific operation: The server organizes the candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[0752] Step 5:

[0753] The terminal presents the candidate list to the user.

[0754] Input: Candidate list data sent from the server.

[0755] Output: The candidate list displayed in the chat interface.

[0756] Specific operation: The device receives the candidate list and displays it to the user in the chat interface. For example, "We recommend the following three red wines: 1. Wine A (2500 yen) 2. Wine B (2800 yen) 3. Wine C (3000 yen)."

[0757] Step 6:

[0758] The user inputs feedback on the candidate list.

[0759] Input: User-provided feedback through a chat interface. For example, "Do you have any wines that are a little sweeter?"

[0760] Output: Feedback is sent as text data from the device to the server.

[0761] Specific operation: The device receives the user's feedback and sends it to the server as an API request.

[0762] Step 7:

[0763] The server receives the feedback and passes it back to the generative AI model in the form of a prompt sentence.

[0764] Input: Feedback text data sent from the device.

[0765] Output: The feedback converted into a new prompt format is passed to the generative AI model.

[0766] Specific operation: The server converts the feedback text into a prompt sentence format and passes it to the generative AI model as "Please suggest a sweeter red wine."

[0767] Step 8:

[0768] The generative AI model then generates new food and drink suggestions based on the new criteria.

[0769] Input: The new prompt text passed by the server.

[0770] Output: A list of new candidate foods and drinks (wine).

[0771] Specific operation: The generative AI model selects sweet red wines from the database based on the new criteria and generates a candidate list. For example, it selects "Wine D (2,600 yen)" and "Wine E (2,900 yen)."

[0772] Step 9:

[0773] The server organizes a new candidate list and sends it to the terminal again.

[0774] Input: A new candidate list generated by a generative AI model.

[0775] Output: The new, cleaned candidate list is sent to the terminal.

[0776] Specific operation: The server organizes the new candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[0777] Step 10:

[0778] The terminal presents the new candidate list to the user.

[0779] Input: The new candidate list data sent from the server.

[0780] Output: The new candidate list displayed in the chat interface.

[0781] Specific operation: The device displays the new candidate list received by the device to the user in the chat interface. For example, "Recommended sweet red wines are Wine D (2600 yen) and Wine E (2900 yen)."

[0782] Step 11:

[0783] The user selects the best wine.

[0784] Input: Final candidates selected by the user through the chat interface.

[0785] Output: The selection results are sent as text data from the terminal to the server.

[0786] Specific operation: The user enters "I will choose wine E," and the device sends the selection result to the server as an API request.

[0787] Step 12:

[0788] The terminal displays a purchase page to the user.

[0789] Input: Purchase page generation instructions from the server.

[0790] Output: Purchase page displayed on the information device.

[0791] Specific operation: The server generates a link to the purchase page and sends it to the device, and the device opens the link and displays the purchase page.

[0792] Step 13:

[0793] The user completes the purchase.

[0794] Input: User's payment information on the purchase page.

[0795] Output: Confirmation that the purchase was successful.

[0796] What happens: The user enters payment information on the purchase page and completes the checkout process.

[0797] Step 14:

[0798] The server processes the payment and arranges shipping.

[0799] Input: Checkout and payment information.

[0800] Output: Confirmation that payment processing has been completed and shipping arrangements have been made.

[0801] Specific operation: The server works with the payment gateway to process the payment and uses the logistics API to arrange for shipping.

[0802] (Application example 1)

[0803] 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."

[0804] Conventional wine selection systems are limited to specific locations, such as the user's home or office. Furthermore, their reliance on smartphones or tablets limits the real-time in-store purchasing experience. Furthermore, while immediate feedback from users is required, there is a lack of efficient means to achieve this. Therefore, there is a need for a system that allows users to efficiently select the optimal wine in-store.

[0805] 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.

[0806] In this invention, the server includes means for receiving conditions input by a user in a chat format using a terminal, a generative AI model for generating candidate wines based on the received conditions, means for displaying a list of candidate wines on the smart glasses, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for displaying the list of re-suggested candidate wines on the smart glasses, and means for generating a procedure page for purchasing the wine selected by the user. This allows the user to select the optimal wine in real time in the store and quickly receive re-suggested wines based on the feedback.

[0807] "User" means an individual or group that uses a service or product.

[0808] A "terminal" is an electronic device used by a user to input conditions, such as a smartphone, tablet, PC, or smart glasses.

[0809] "Chat style" refers to an interface style in which users exchange text and voice in a natural conversational flow.

[0810] "Conditions" refer to specific requests or requirements entered by the user (e.g., budget, preferred flavor, usage scenario, etc.).

[0811] "Receiving" refers to the act of receiving input data or information.

[0812] A "generative AI model" is an artificial intelligence model that is trained to generate optimal candidates based on a user's criteria.

[0813] A "candidate list" is a list of multiple options selected by a generative AI model.

[0814] "Feedback" refers to opinions and additional requests provided by users.

[0815] "Re-proposal" refers to the act of presenting a newly selected candidate based on the feedback received.

[0816] The "procedure page" is a web page where the user inputs and confirms the information required to purchase the selected wine.

[0817] "Smart glasses" are electronic devices in the form of glasses that have a built-in display and camera and have the ability to display information superimposed on the user's field of vision.

[0818] "Display" refers to the act of visually presenting information on the screen of a device or smart glasses.

[0819] "Purchase" refers to the act of obtaining goods or services in exchange for money.

[0820] This invention provides a wine concierge system for efficiently selecting wines that meet the user's needs. Specifically, the user inputs criteria using smart glasses, and a generative AI model generates an optimal candidate list. Based on the feedback, a new candidate list of wines is then displayed, allowing the user to select the optimal wine and proceed with the purchase.

[0821] System Configuration

[0822] The system consists of the following main components:

[0823] 1. User Device:

[0824] A device that allows users to input conditions, including smart glasses. Smart glasses display a chat-style interface and allow voice or touch input. Google Glass is a typical example.

[0825] 2. Server:

[0826] It receives conditions sent from the user's device and passes them to the generative AI model. The server runs in a cloud environment (e.g., an EC2 instance on AWS). Here, it is built as a web application using Flask.

[0827] 3. Generative AI Model:

[0828] It is a generative AI model trained using PyTorch to generate optimal wine candidates based on user input criteria. Specifically, it uses the GPT-3 model.

[0829] Program processing

[0830] The server receives conditional input from the user using a Flask application.

[0831] The received criteria are passed to the GPT-3 model to generate an optimal wine candidate list.

[0832] The server sends the generated wine candidate list to the smart glasses, which then overlays it on the field of view.

[0833] When feedback is received, the conditions are passed to the generative AI model again, and a regenerated candidate list is displayed on the smart glasses.

[0834] Finally, a procedure page for purchasing the wine selected by the user is generated and displayed on the smart glasses.

[0835] Specific examples

[0836] When a user wants to select wine for a house party, the following steps occur:

[0837] 1. Enter the conditions:

[0838] The user voice-inputs into the chat interface of the smart glasses, "I'm looking for wine to use at a home party. My budget is under 4,000 yen, and I'd prefer a white wine."

[0839] 2. Generate wine candidates:

[0840] The server passes the received criteria to the GPT-3 model, which generates an optimal wine candidate list based on the criteria.

[0841] 3. Displaying the results:

[0842] A list of wine candidates is sent from the server to the smart glasses and displayed over the user's field of view. For example, it might look like this: "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[0843] 4. Providing Feedback:

[0844] When the user voices feedback such as "Do you have any sweeter wines?", the server receives this and passes the new criteria to the generative AI model again to generate a revised list of candidates.

[0845] 5. View resubmissions:

[0846] The regenerated candidate list is displayed on the smart glasses. For example, it may look like this: "Candidate 1: Wine C (3,600 yen, sweet)."

[0847] 6. Purchase Procedure:

[0848] The user selects the most suitable wine and inputs the selection result by saying, "I will choose wine C." Based on this, the server generates a checkout page and displays it on the smart glasses.

[0849] Prompt Sentence Examples

[0850] As a concrete example, the prompt sentence "Home party wine budget 4000 yen white wine" is passed to the generative AI model. Based on this prompt sentence, the generative AI model will suggest the most suitable wine candidates.

[0851] The system allows users to use smart glasses to select the perfect wine in real time and quickly receive additional suggestions based on feedback.

[0852] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0853] Step 1:

[0854] The user inputs the requirements by voice into the chat interface of the smart glasses. In this step, the voice is converted into text using a speech recognition API and stored in a database. The specific input is "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable." The output is the textual requirements: "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable."

[0855] Step 2:

[0856] The device sends the textual conditions to the cloud server, which then receives the conditions and prepares them for passing to the generative AI model. The input is the textual conditions, and the output is an HTTP request containing the conditions.

[0857] Step 3:

[0858] The server passes the received conditions to a generative AI model to generate a list of wine candidates. Specifically, it uses a GPT-3 model to generate the best wine candidates based on the conditions. The input is text conditions, and the conditions are converted into prompt sentences as data processing and input into the generative AI model. The output is a list of wine candidates. For example, "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[0859] Step 4:

[0860] The server sends the generated wine candidate list to the smart glasses and displays it over the user's field of view. In this step, data is sent via an HTTP response and the information is displayed on the smart glasses' display. The input is the wine candidate list, and the output is the information overlaid on the smart glasses' display.

[0861] Step 5:

[0862] The user speaks their feedback to the displayed list of candidates, and the feedback is converted into text. For example, they can say, "Do you have any sweeter wines?" and the speech recognition API is used to convert it into text. The input is spoken feedback, and the output is text feedback.

[0863] Step 6:

[0864] The device sends textual feedback to the server, which receives and analyzes it. The conditions are passed to the generative AI model again to generate a new candidate list. The input is the textual feedback, which is converted into a new prompt sentence as data processing and input to the generative AI model. The output is a regenerated list of wine candidates. For example, "Candidate 1: Wine C (3,600 yen, sweet)."

[0865] Step 7:

[0866] The server sends the regenerated wine candidate list to the smart glasses and displays it over the user's field of view. The input is the regenerated wine candidate list, and the output is the information overlaid on the smart glasses' display.

[0867] Step 8:

[0868] The user selects the most suitable wine and inputs the selection by voice or touch. For example, "I will choose wine C." The input is the user's selection, and the output is a text version of the selection.

[0869] Step 9:

[0870] The terminal sends the selection result to the server, and the server generates a checkout page. In this step, a web page containing the information necessary for the checkout process is generated based on the selection result and sent to the smart glasses. The input is the selection result, and the output is the checkout page.

[0871] Step 10:

[0872] The user enters the necessary information on the checkout page displayed on the smart glasses to complete the purchase. Specifically, they enter credit card information and shipping address information. The input is the user's purchase procedure information, and the output is a confirmation message that the purchase has been completed.

[0873] Through these processing steps, users can select the best wine in real time and quickly receive additional suggestions based on feedback.

[0874] 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.

[0875] This invention combines a wine concierge system that utilizes a generative AI model that suggests the most suitable wine based on user input criteria such as budget and preferences when selecting a wine, with an emotion engine that recognizes the user's emotions. Below, the program processing of this system is explained in natural language, with specific examples.

[0876] System Overview

[0877] The system is configured as follows:

[0878] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[0879] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred flavor, usage scenario, etc.).

[0880] 3. The terminal sends the entered conditions to the server as text data.

[0881] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[0882] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[0883] 6. The generative AI model searches a database of wines based on criteria and sentiment to generate the best wine candidates.

[0884] 7. The server organizes the generated candidate list and sends it to the terminal.

[0885] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[0886] 9. The user enters feedback on the presented wine candidates (e.g., "I'd like something sweeter" or "I'm willing to increase my budget")

[0887] 10. The device sends the feedback to the server.

[0888] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[0889] 12. The generative AI model again generates wine suggestions based on criteria and sentiment.

[0890] 13. The server sends the re-proposal candidate list to the terminal.

[0891] 14. The device again presents the chat interface to the user.

[0892] 15. The device will then display a page where the user can select the perfect wine and complete the purchase.

[0893] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[0894] Examples:

[0895] Example 1: Choosing wine for a home party

[0896] 1. The user enters the criteria

[0897] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[0898] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[0899] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[0900] The server receives the conditions and emotion data and passes them to the generative AI model.

[0901] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[0902] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[0903] 3. Users provide feedback

[0904] User: "Do you have any sweeter wines?"

[0905] The device sends this feedback and the user's emotional data to the server.

[0906] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[0907] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[0908] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[0909] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[0910] 5. The user makes the final decision and purchases the wine.

[0911] User: "I'll choose wine E."

[0912] The terminal transmits the selection result to the server.

[0913] The server generates a purchase page and sends it to the terminal.

[0914] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[0915] This allows users to find the perfect wine that reflects their emotions. Through this process, users can significantly reduce the effort and time required to select a wine while also improving their emotional satisfaction.

[0916] The processing flow will be explained below.

[0917] Step 1:

[0918] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[0919] Step 2:

[0920] The user enters the conditions into a chat-style interface. For example, the user might enter, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device then sends this condition to the server as text data.

[0921] Step 3:

[0922] The server receives the conditions. At the same time, the emotion engine analyzes the emotional data from the user's input and generates emotional information such as "joy" or "expectation." This data is then passed to the generative AI model by the server.

[0923] Step 4:

[0924] A generative AI model searches a database of wines based on criteria (budget, red wine) and emotional data (e.g., joy, anticipation). The AI ​​model generates optimal wine candidates and sends them back to the server.

[0925] Step 5:

[0926] The server organizes the wine recommendation list received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, and emotional reasons for the recommendation.

[0927] Step 6:

[0928] The terminal displays the formatted wine candidate list on the chat interface and presents it to the user, who can then confirm the presented candidate list.

[0929] Step 7:

[0930] The user inputs feedback for the presented wine candidates. For example, the user might input, "Do you have any wines that are a little sweeter?" The device then sends this feedback to the server.

[0931] Step 8:

[0932] The server receives the user feedback and analyzes the emotion data again through the emotion engine, which prepares to pass the updated emotion information to the generative AI model.

[0933] Step 9:

[0934] The server passes the feedback and updated emotional data to the generative AI model, which then re-searches for wine candidates based on the new criteria (sweet red wine) and the updated emotional data.

[0935] Step 10:

[0936] The generative AI model generates new wine candidates based on the new criteria and emotional data and sends them back to the server, which receives them and organizes the list of candidates for re-suggestion.

[0937] Step 11:

[0938] The terminal displays the re-suggested wine candidate list in the chat interface and presents it to the user. The re-suggested candidate list includes information based on the new conditions and emotional information.

[0939] Step 12:

[0940] The user selects the best wine. For example, the user inputs, "I will choose wine E." The device sends this selection result to the server.

[0941] Step 13:

[0942] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[0943] Step 14:

[0944] The terminal displays a purchase procedure page to the user using a chat interface, where the user enters the necessary information and completes the purchase procedure.

[0945] Step 15:

[0946] Once the user completes the purchase, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[0947] Step 16:

[0948] The device will display a notification to the user that the purchase is complete and provide information such as the estimated delivery date, allowing the user to select wine efficiently and emotionally in a satisfying way.

[0949] Example 2

[0950] 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."

[0951] Conventional wine selection systems only consider the criteria entered by the user to suggest products, which tends to result in suggestions that ignore the user's feelings. This reduces user satisfaction and makes it difficult to select the appropriate product. In addition, the process of re-suggesting products based on feedback often does not function efficiently, resulting in a suboptimal user experience.

[0952] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing emotions from user input data, means for transferring the analyzed emotion data to the information processing model, and means for analyzing feedback provided by the user and transferring conditions based on the feedback back to the information processing model. This makes it possible to propose products that reflect the user's emotions, improving user satisfaction and realizing an efficient feedback re-proposal process.

[0953] A "user" is a user who uses a terminal to access the service and input conditions and feedback.

[0954] "Terminal" refers to an electronic device used by a user, such as a smartphone, tablet, or PC, and is a device for inputting conditions and feedback through a chat-style interface.

[0955] The "server" is a central system that processes data received from users, performs sentiment analysis, product suggestions and re-suggestions, and sends the results to the terminal.

[0956] The "conditions" refer to the user's wishes and requirements regarding the product he or she wishes to purchase, and specifically include budget, preferred type, purpose, and the like.

[0957] An "information processing model" is an algorithm or system that generates optimal product candidates based on user-entered conditions and sentiment analysis data.

[0958] A "generative AI model" is a type of information processing model, a system that uses AI technology to generate product candidates based on user conditions and emotional data.

[0959] An "emotion engine" is a system that analyzes emotions from user input data and provides the results to a generative AI model.

[0960] "Feedback" refers to additional opinions or requests provided by the user regarding the proposed product candidates, and serves as reference information when the system makes another suggestion.

[0961] A "product candidate list" is a list of candidate products that are judged to meet the conditions from among the group of products proposed by the generative AI model.

[0962] "Re-proposal" refers to the process in which the generative AI model regenerates and proposes a list of product candidates based on user feedback.

[0963] This invention combines an emotion engine that recognizes the user's emotions with a product concierge system that utilizes an information processing model that suggests optimal products based on user input criteria such as budget and preferences when selecting a product. The program processing of this system is explained below, along with specific examples.

[0964] System Overview

[0965] The system is configured as follows:

[0966] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[0967] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred features, usage scenario, etc.).

[0968] 3. The terminal sends the entered conditions to the server as text data.

[0969] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[0970] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[0971] 6. The generative AI model searches the product database based on the conditions and sentiment to generate the best product candidates.

[0972] 7. The server organizes the generated candidate list and sends it to the terminal.

[0973] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[0974] 9. The user enters feedback on the presented product candidates (e.g., "I'd like something sweeter" or "I'd be okay with a higher budget").

[0975] 10. The device sends the feedback to the server.

[0976] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[0977] 12. The generative AI model again generates product candidates based on the conditions and sentiment.

[0978] 13. The server sends the re-proposal candidate list to the terminal.

[0979] 14. The device again presents the chat interface to the user.

[0980] 15. A page will be displayed on the device that allows the user to select the most suitable product and complete the purchase process.

[0981] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[0982] Hardware and software used

[0983] Users use devices such as smartphones, tablets, and PCs.

[0984] The servers are high-performance cloud servers equipped with data processing and storage capabilities.

[0985] The emotion engine uses natural language processing technology to analyze emotions from user input data.

[0986] Generative AI models use machine learning algorithms to suggest the best products based on criteria and sentiment.

[0987] Specific examples

[0988] Example 1: Choosing wine for a home party

[0989] 1. The user enters the criteria

[0990] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[0991] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[0992] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[0993] The server receives the conditions and emotion data and passes them to the generative AI model.

[0994] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[0995] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)" is sent to the terminal.

[0996] 3. Users provide feedback

[0997] User: "Do you have any sweeter wines?"

[0998] The device sends this feedback and the user's emotional data to the server.

[0999] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[1000] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[1001] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[1002] Server: Sends the message "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)" to the terminal.

[1003] 5. The user makes the final decision and purchases the wine.

[1004] User: "I'll choose wine E."

[1005] The terminal transmits the selection result to the server.

[1006] The server generates a purchase page and sends it to the terminal.

[1007] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[1008] This entire process allows users to find the perfect product that reflects their emotions and complete the purchase process smoothly.

[1009] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1010] Step 1:

[1011] The user enters the conditions

[1012] Input: Budget, preferred features, usage scenario, etc.

[1013] Output: Text data containing the condition

[1014] Specific operation: The user opens a chat interface on the device and enters criteria in text format, such as "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device captures this input as text data and immediately sends it to the server.

[1015] Step 2:

[1016] The device sends the user-entered data to the server

[1017] Input: Text data of the condition entered by the user

[1018] Output: Text data sent to the server

[1019] Specific operation: The terminal sends the user's input data to the server via secure communication. The input data is converted into the required format and then passed to the server.

[1020] Step 3:

[1021] The server analyzes the condition and emotion

[1022] Input: Condition data received from the user

[1023] Output: Condition data and emotion data

[1024] Specific operation: The server passes the received text data to the emotion engine, which analyzes the emotions (joy, anticipation, etc.) from the user's input. Condition data and emotion data are generated as the analysis results.

[1025] Step 4:

[1026] The server passes the analysis results to the generative AI model

[1027] Input: Condition data and emotion data

[1028] Output: The data fed into the generative AI model

[1029] Specific operation: The server integrates the emotion data obtained by the emotion engine with the user's condition data and passes it to the generative AI model.

[1030] Step 5:

[1031] Generative AI model generates product candidates

[1032] Input: Condition data and emotion data

[1033] Output: Product candidate list

[1034] Specific operation: The generative AI model searches the database based on the conditions and emotions. For example, if the conditions are "red wine, budget under 3,000 yen," it retrieves appropriate products (e.g., Wine A, Wine B, Wine C) from the database and generates a product candidate list.

[1035] Step 6:

[1036] The server sends the generated candidate list to the device.

[1037] Input: Product candidate list

[1038] Output: Product candidate list sent to the terminal

[1039] Specific operation: The server organizes the generated candidate list, converts it into an easy-to-understand format, and then sends the candidate list to the terminal.

[1040] Step 7:

[1041] The device presents a candidate list to the user.

[1042] Input: Product candidate list sent from the server

[1043] Output: Product candidate list displayed in chat interface

[1044] Specific operation: The device presents the user with a list of options via a chat-style interface, such as "We recommend the following three products: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)."

[1045] Step 8:

[1046] User enters feedback

[1047] Input: Feedback on product candidate list

[1048] Output: Feedback text data

[1049] Specific operation: The user inputs feedback such as "Do you have any wines that are a little sweeter?", and the device captures this feedback as text data and sends it to the server.

[1050] Step 9:

[1051] The device sends feedback to the server

[1052] Input: User feedback text data

[1053] Output: Feedback data sent to the server

[1054] How it works: The device sends feedback to the server over secure communication, and the data is converted into the required format and then passed to the server.

[1055] Step 10:

[1056] The server analyzes the emotions and feedback again.

[1057] Input: Feedback data

[1058] Output: Updated condition data and emotion data

[1059] Specific operation: The server passes the feedback to the emotion engine, which analyzes the emotion again and updates the condition data based on the feedback. For example, if the user's feedback is "sweeter," the condition data is updated to "sweet red wine."

[1060] Step 11:

[1061] The generative AI model generates new product candidates

[1062] Input: Updated condition data and emotion data

[1063] Output: Regenerated product candidate list

[1064] Specific operation: The generative AI model re-searches the database based on new criteria (e.g., sweet red wine) and emotion data to generate new product candidates. If the new list is "Wine D (2,600 yen), Wine E (2,900 yen)," etc., it creates that list.

[1065] Step 12:

[1066] The server sends a list of re-proposals to the device.

[1067] Input: Regenerated product candidate list

[1068] Output: Regeneration candidate list sent to the terminal

[1069] Specific operation: The server organizes the regenerated candidate list, converts it into an easy-to-understand format, and sends it to the terminal.

[1070] Step 13:

[1071] The device presents the candidate list to the user again.

[1072] Input: Regeneration candidate list sent from the server

[1073] Output: Regeneration candidate list displayed in the chat interface

[1074] Specific operation: The device presents the user with a list of regeneration candidates via a chat-style interface, such as "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[1075] Step 14:

[1076] The user decides on the best product and purchases it.

[1077] Input: Final selected product

[1078] Output: Purchase procedure page generated and payment and shipping arrangements completed

[1079] Specific operation: The user enters "I'll choose wine E," and the device sends the selection result to the server. The server generates a purchase procedure page and sends it to the device. The user completes the purchase procedure, and payment and shipping arrangements are completed.

[1080] (Application example 2)

[1081] 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."

[1082] Conventional wine recommendation systems have been unable to select wines that reflect the user's emotions, making it difficult to recommend the best wine for the user. Furthermore, when selecting wine in a brick-and-mortar store, there is a lack of a way for users to efficiently input their budget and preferences and obtain the best wine in real time. The present invention aims to solve these problems and provide a system that can make better wine recommendations that reflect the user's emotions.

[1083] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving conditions entered by a user in a chat format using a terminal; a generative AI model that generates candidate wines based on the received conditions and the user's emotional data; means for transmitting the generated candidate wine list to the terminal; means for receiving feedback provided by the user and again passing the conditions to the generative AI model; means for transmitting the re-proposed candidate wine list to the terminal; means for generating a procedure page for purchasing the wine selected by the user; means for recognizing the user's emotions; means for transferring the emotional data to the generative AI model; and means for using a display device to present the wine selection results to the user. This enables optimal wine recommendations that reflect the user's emotions, enabling efficient wine selection even in physical stores.

[1084] "Means for receiving conditions entered by a user in chat format using a terminal" refers to a mechanism by which a user enters wine selection conditions in chat format using a device such as a smartphone or tablet, and receives that input.

[1085] A "generative AI model that generates candidate wines based on received conditions and user emotional data" is an artificial intelligence model that generates optimal wine candidates based on conditions and emotional data entered by the user.

[1086] The "means for transmitting the generated wine candidate list to the terminal" is a communication function for transmitting the wine candidate list generated by the AI ​​model to the user's device.

[1087] "Means for receiving feedback provided by the user and passing conditions back to the generative AI model" refers to a mechanism for receiving feedback from the user and providing conditions back to the generative AI model based on that feedback.

[1088] The "means for sending a re-proposed wine candidate list to the terminal" is a function for sending a re-generated wine candidate list generated by the AI ​​model based on the updated conditions to the user's device.

[1089] The "means for generating a procedure page for purchasing the wine selected by the user" is a system for generating a web page for purchasing the wine finally selected by the user.

[1090] "Means for recognizing user emotions" refers to technology that analyzes the user's voice, facial expressions, etc. to identify their emotions.

[1091] "Means for transferring emotion data to the generative AI model" refers to a function for transmitting emotion data obtained by emotion recognition to the generative AI model.

[1092] The "means of using a display device to present wine selection results to a user" refers to a technology that uses smart glasses or other display devices to display wine selection results to a user.

[1093] A system for implementing the present invention uses the following hardware and software.

[1094] Hardware:

[1095] Smart devices (smartphones, tablets)

[1096] Smart glasses (e.g., Google Glass)

[1097] server

[1098] software:

[1099] Python

[1100] Flask (web server framework)

[1101] OpenAI API (generative AI model)

[1102] Emotion analysis software (e.g., voice and facial expression recognition libraries)

[1103] System Overview

[1104] The system is configured as follows:

[1105] 1. User input:

[1106] Using a smart device or smart glasses, users can input wine selection criteria (e.g., budget, preferred flavor, intended use, etc.) in a chat format. The device then sends these criteria as data to the server.

[1107] 2. Collecting Emotional Data:

[1108] The smart glasses analyze the user's voice and facial expressions in real time to collect emotional data, which is then sent to the server along with the condition data.

[1109] 3. Sentiment Analysis and Conditional Processing:

[1110] The server analyzes the received conditions and emotion data and uses emotion analysis software to understand the user's emotions.

[1111] 4. Wine selection using generative AI models:

[1112] The server inputs a prompt sentence into the generative AI model based on the emotion data and condition data, and generates wine candidates. For example, the prompt sentence might look like this:

[1113] User sentiment: positive

[1114] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[1115] Choose the three best wines.

[1116] 5. User suggestions for wines:

[1117] The server generates a list of wine candidates, which is then sent to the smart device or smart glasses and displayed to the user. For example, it might show "Wine A (2,500 yen), Wine B (2,800 yen), Wine C (3,000 yen)."

[1118] 6. Feedback and Resubmissions:

[1119] The user inputs feedback on the presented wine candidates (e.g., "A slightly sweeter wine"), and the device sends this feedback and the user's emotional data to the server. The server then performs emotional analysis again, passes new conditions to the generative AI model, and generates a new list of wine candidates, which it then presents to the user.

[1120] 7. Final Decision and Purchase Procedure:

[1121] Once the user selects the wine that best suits them, the device sends the selection results to the server, which then generates a purchase procedure page and displays it to the user. The user completes the purchase procedure on that page, and payment and shipping arrangements are completed.

[1122] Examples:

[1123] Wine selection for a home party:

[1124] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[1125] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[1126] User: "Do you have any sweeter wines?"

[1127] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[1128] User: "I'll choose wine E."

[1129] This system allows users to find the perfect wine that reflects their own feelings, and also streamlines the selection process in physical stores.

[1130] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1131] Step 1:

[1132] The user uses a smart device or smart glasses to input wine selection criteria (e.g., budget, preferred taste, intended use, etc.) in chat format. This generates input data (criteria) from the user.

[1133] Step 2:

[1134] The terminal receives the data entered by the user and sends it to the server. At this time, the data is sent in text format. An example of data sent from the terminal to the server is "Budget: under 3000 yen, red wine for a home party."

[1135] Step 3:

[1136] The smart glasses analyze the user's voice and facial expressions in real time to generate emotional data. The emotional data is analyzed as expressions such as "happiness" or "expectation" using voice and facial recognition software. The generated emotional data is then sent back to the server.

[1137] Step 4:

[1138] The server analyzes the received condition data and emotion data and generates an analysis result using emotion analysis software. Specifically, a voice analysis engine and a facial expression analysis engine are used. Based on this, the user's emotional state is determined to be "positive."

[1139] Step 5:

[1140] The server combines the condition data and emotion data to generate a prompt sentence, which is then input to the generative AI model. An example of this prompt sentence is as follows:

[1141] User sentiment: positive

[1142] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[1143] Choose the three best wines.

[1144] The generative AI model generates a list of optimal wines based on this prompt.

[1145] Step 6:

[1146] The server receives the wine candidate list output from the generative AI model and sends it to the terminal. An example of a wine candidate list that is sent is "Wine A (2500 yen), Wine B (2800 yen), Wine C (3000 yen)."

[1147] Step 7:

[1148] The device receives the wine candidate list and presents it to the user in a chat format. The user then checks the candidate list displayed on the display of the smart glasses or smart device.

[1149] Step 8:

[1150] The user inputs feedback on the wine candidate list, for example, "Do you have any sweeter wines?" This feedback is sent to the server via the terminal.

[1151] Step 9:

[1152] The server analyzes the feedback data and the new emotion data, and again inputs new condition prompts into the generative AI model, which generates a revised wine candidate list. For example, the following prompts are used:

[1153] User sentiment: positive

[1154] Wine selection criteria: A sweet red wine for a home party, within a budget of 3,000 yen

[1155] Choose the three best wines.

[1156] Step 10:

[1157] The server receives the re-proposed wine candidate list and sends it to the terminal. An example of the re-proposed list sent is "Wine D (2600 yen), Wine E (2900 yen)".

[1158] Step 11:

[1159] The terminal again presents the user with a list of wine candidates, and the user selects the most suitable wine. The user then makes a final selection, such as "I will choose wine E."

[1160] Step 12:

[1161] The server generates a procedure page for purchasing the wine selected by the user and sends it to the terminal. The user completes the purchase procedure on the page, and once payment is complete, shipping arrangements are made.

[1162] 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.

[1163] 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.

[1164] 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.

[1165] [Third embodiment]

[1166] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1167] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1168] 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).

[1169] 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.

[1170] 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.

[1171] 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).

[1172] 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.

[1173] 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.

[1174] 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.

[1175] 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.

[1176] 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.

[1177] 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."

[1178] This invention is a wine concierge system that utilizes a generative AI model to suggest the most suitable wine based on the user's input criteria such as budget and preferences when selecting a wine. Below, we will explain the program processing of this system in natural language and provide specific examples.

[1179] System Overview

[1180] The system is configured as follows:

[1181] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[1182] 2. The device displays a chat-style interface to the user, and the user enters their requirements (budget, preferred flavor, usage situation, etc.).

[1183] 3. The terminal sends the entered conditions to the server.

[1184] 4. The server receives the conditions and passes the data to the generative AI model.

[1185] 5. The generative AI model generates optimal wine candidates based on the criteria and sends them back to the server.

[1186] 6. The server organizes the generated candidate list and sends it to the terminal.

[1187] 7. The device displays the candidate list on the chat interface and presents it to the user.

[1188] 8. The user enters feedback on the presented options (e.g., "I'd like something a little sweeter" or "I'd be happy to increase the budget").

[1189] 9. The device sends feedback to the server, which again passes the conditions to the generative AI model.

[1190] 10. The generative AI model generates wine candidates again based on the new criteria.

[1191] 11. The server sends the new candidate list to the terminal, which again presents it to the user in the chat interface.

[1192] 12. The device will then display a page where the user can select the perfect wine and complete the purchase.

[1193] 13. Once the user completes the purchase, the server processes the payment and arranges for shipping.

[1194] Examples:

[1195] Example 1: Choosing wine for a home party

[1196] 1. The user enters the criteria

[1197] The user enters, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[1198] The terminal transmits this condition to the server as text data.

[1199] 2. The server receives the conditions and the AI ​​model generates candidates.

[1200] The server receives the conditions and passes them to the generative AI model.

[1201] The generative AI model selects three red wines from a database that meet the criteria.

[1202] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[1203] 3. Users provide feedback

[1204] User: "Do you have any sweeter wines?"

[1205] The terminal sends this feedback to the server.

[1206] 4. The server receives the feedback and the AI ​​model generates candidates again.

[1207] The server receives the feedback and again passes the new conditions to the generative AI model.

[1208] A generative AI model selects a sweet red wine.

[1209] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[1210] 5. The user makes the final decision and purchases the wine.

[1211] User: "I'll choose wine E."

[1212] The terminal transmits the selection result to the server.

[1213] The server generates a purchase page and sends it to the terminal.

[1214] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[1215] This allows users to efficiently and effectively find the wine that suits them, and through this process, users can significantly reduce the effort and time required to select a wine.

[1216] The processing flow will be explained below.

[1217] Step 1:

[1218] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[1219] Step 2:

[1220] The user inputs criteria into a chat-style interface. These criteria include "budget," "preferred flavor," and "usage occasion." The user might input something like, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[1221] Step 3:

[1222] The device sends the entered conditions as text data to the server, which then analyzes the data, converts it into an appropriate format, and prepares it for passing to the generative AI model.

[1223] Step 4:

[1224] The server receives the criteria sent by the user and passes them to the generative AI model, which searches a wine database based on the criteria and generates optimal wine candidates.

[1225] Step 5:

[1226] The generative AI model selects several wine candidates that meet the criteria, for example, three highly rated red wines within a budget of 3,000 yen.

[1227] Step 6:

[1228] The server organizes the candidate list of wines received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, etc.

[1229] Step 7:

[1230] The server sends the formatted wine candidate list to the terminal, which displays the received data on the chat screen and presents wine candidates to the user.

[1231] Step 8:

[1232] The user can input feedback on the presented wine candidates, for example, specifying additional conditions such as "I would prefer something a little sweeter" or "I'm willing to increase my budget a little."

[1233] Step 9:

[1234] The device sends the user's feedback as text data to the server, which then prepares to pass the data to the generative AI model again based on the new conditions.

[1235] Step 10:

[1236] The server receives the feedback from the user, analyzes it, and passes the new criteria to the generative AI model, which then re-searches for wine candidates based on the new criteria.

[1237] Step 11:

[1238] The generative AI model again selects wine candidates that meet the new criteria and sends them back to the server, which then organizes the list of re-suggested wines.

[1239] Step 12:

[1240] The server sends the reformatted list of wine candidates to the terminal, which then displays the received data on the chat screen and re-proposes the wine to the user.

[1241] Step 13:

[1242] The user selects the best wine from the provided candidates and sends the selection result to the server via the terminal.

[1243] Step 14:

[1244] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[1245] Step 15:

[1246] The terminal displays a checkout page to the user in a chat interface, where the user can complete the purchase.

[1247] Step 16:

[1248] Once the user completes the purchase procedure, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[1249] Step 17:

[1250] The device will notify the user of the completion of the purchase and provide information such as the estimated delivery date, allowing the user to select and purchase wine efficiently and effectively.

[1251] Example 1

[1252] 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."

[1253] Conventional wine selection systems make it difficult for users to effectively select wines that match their preferences and requirements, and the selection process takes a lot of time and effort. Furthermore, the process of reflecting user feedback and re-recommending wines is complicated and inefficient. Another problem is that purchasing procedures, payment processing, and shipping arrangements cannot be managed in a single system.

[1254] 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.

[1255] In this invention, the server includes means for receiving conditions input by a user in chat format using an information device, a generative AI model that generates candidate foods and beverages based on the received conditions, means for transmitting the generated candidate food and beverage list to the information device, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for transmitting the re-proposed candidate food and beverage list to the information device, means for generating a procedure page for purchasing the food and beverage selected by the user, means for passing the conditions received from the terminal to the generative AI model in the form of a prompt sentence, means for passing the feedback again to the generative AI model in the form of a prompt sentence when the user provides new feedback, and means for processing payment and arranging for delivery to complete the purchase procedure. This allows users to efficiently and effectively find the food and beverages that suit them and easily complete the purchase procedure.

[1256] An "information device" is a terminal that a user can access (such as a smartphone, tablet, or PC).

[1257] A "generative AI model" is an artificial intelligence algorithm that generates candidate foods and beverages based on received conditions.

[1258] A "prompt format" is a textual description of a question or instruction provided to a generative AI model.

[1259] A "candidate list" is a list of multiple foods and drinks suggested by the generative AI model.

[1260] "Feedback" refers to a reaction, including hopes, requests, corrections, etc., provided by a user to a candidate list.

[1261] "Payment processing" refers to the procedure for completing the payment required to purchase the food and drink selected by the user.

[1262] "Shipping arrangements" are logistics procedures for delivering purchased food and drink to the user.

[1263] "Chat style" refers to an interface between the user and the system that is based on the exchange of text messages.

[1264] "Conditions" are selection criteria entered by the user, such as budget, preferred flavor, and usage scenario.

[1265] The "procedure page" is a web page or app screen for the user to purchase the food and drink selected by the user.

[1266] The present invention is a concierge system that utilizes a generative AI model to suggest optimal food and drink (e.g., wine) based on user input criteria. This system is implemented primarily using information devices, a server, and a generative AI model. Specific embodiments of this system are described below.

[1267] System configuration

[1268] The system is configured as follows:

[1269] 1. Information devices (terminals):

[1270] Devices such as smartphones, tablets, and computers are used.

[1271] These terminals display a chat-style interface for users to enter requirements and provide feedback.

[1272] 2. Server:

[1273] The server implements a generative AI model.

[1274] It receives conditions and feedback entered by the user through the device and passes them on to the generative AI model.

[1275] The generated candidate list is also organized and transmitted to the terminal.

[1276] 3. Generative AI Model:

[1277] Generative AI models (e.g., OpenAI GPT-4 or Transformer-based models) generate optimal food and drink candidates based on criteria and feedback provided by the server.

[1278] The model selects suitable candidates from the database and returns this information to the server.

[1279] System operation example

[1280] As a concrete example, let us consider a scenario of selecting wine for a home party.

[1281] 1. User enters criteria:

[1282] A user enters the following into the chat interface of an information device:

[1283] "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[1284] 2. The server passes the conditions to the generative AI model:

[1285] The server receives this condition and sends it to the generative AI model in the form of a prompt sentence.

[1286] For example: "Please suggest a red wine for a home party within the budget of 3000 yen."

[1287] 3. The generative AI model generates wine candidates:

[1288] Based on the criteria, the generative AI model generates wine candidates such as:

[1289] "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[1290] 4. Users provide feedback:

[1291] The user enters feedback.

[1292] For example: "Do you have any wines that are a little sweeter?"

[1293] 5. The server passes the feedback back to the generative AI model:

[1294] The server then sends the received feedback back to the generative AI model in the form of a prompt sentence.

[1295] For example: "Please suggest a sweeter red wine."

[1296] 6. The generative AI model again generates wine candidates:

[1297] The generative AI model generates new candidates based on the new conditions, such as:

[1298] "The recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[1299] The user selects the most suitable wine from these options, and the terminal generates a purchase page. Once the purchase procedure is complete, the server processes the payment and arranges for delivery.

[1300] In this way, the present invention not only allows the user to efficiently select food and drink, but also allows the user to complete the purchase procedure in an integrated manner, thereby significantly reducing the time and effort required for selecting food and drink.

[1301] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1302] Step 1:

[1303] The user enters the conditions using a terminal.

[1304] Input: A chat interface displayed on an information device (smartphone, tablet, PC). The user types, "I'm looking for wine to use at a home party. My budget is under 3000 yen, and red wine would be good."

[1305] Output: The conditions are sent as text data from the terminal to the server.

[1306] Specific operation: The device receives user input and sends that data to the server as an API request.

[1307] Step 2:

[1308] The server receives the conditions and passes them to the generative AI model in the form of a prompt sentence.

[1309] Input: Conditional text data sent from the terminal.

[1310] Output: The conditions converted into prompt sentence format are passed to the generative AI model.

[1311] Specific operation: The server converts the condition text into a prompt sentence format and inputs it into the generative AI model. It passes it to the model as "Please suggest a red wine for a home party within a budget of 3,000 yen."

[1312] Step 3:

[1313] A generative AI model generates food and drink suggestions based on criteria.

[1314] Input: The prompt text passed by the server.

[1315] Output: A list of food and drink options (wines).

[1316] Specific operation: The generative AI model selects wines from the database that meet the conditions and generates a candidate list. For example, it selects "Wine A (2,500 yen)," "Wine B (2,800 yen)," and "Wine C (3,000 yen)."

[1317] Step 4:

[1318] The server organizes the generated candidate list and transmits it to the terminal.

[1319] Input: A candidate list generated by a generative AI model.

[1320] Output: The curated candidate list is sent to the terminal.

[1321] Specific operation: The server organizes the candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[1322] Step 5:

[1323] The terminal presents the candidate list to the user.

[1324] Input: Candidate list data sent from the server.

[1325] Output: The candidate list displayed in the chat interface.

[1326] Specific operation: The device receives the candidate list and displays it to the user in the chat interface. For example, "We recommend the following three red wines: 1. Wine A (2500 yen) 2. Wine B (2800 yen) 3. Wine C (3000 yen)."

[1327] Step 6:

[1328] The user inputs feedback on the candidate list.

[1329] Input: User-provided feedback through a chat interface. For example, "Do you have any wines that are a little sweeter?"

[1330] Output: Feedback is sent as text data from the device to the server.

[1331] Specific operation: The device receives the user's feedback and sends it to the server as an API request.

[1332] Step 7:

[1333] The server receives the feedback and passes it back to the generative AI model in the form of a prompt sentence.

[1334] Input: Feedback text data sent from the device.

[1335] Output: The feedback converted into a new prompt format is passed to the generative AI model.

[1336] Specific operation: The server converts the feedback text into a prompt sentence format and passes it to the generative AI model as "Please suggest a sweeter red wine."

[1337] Step 8:

[1338] The generative AI model then generates new food and drink suggestions based on the new criteria.

[1339] Input: The new prompt text passed by the server.

[1340] Output: A list of new candidate foods and drinks (wine).

[1341] Specific operation: The generative AI model selects sweet red wines from the database based on the new criteria and generates a candidate list. For example, it selects "Wine D (2,600 yen)" and "Wine E (2,900 yen)."

[1342] Step 9:

[1343] The server organizes a new candidate list and sends it to the terminal again.

[1344] Input: A new candidate list generated by a generative AI model.

[1345] Output: The new, cleaned candidate list is sent to the terminal.

[1346] Specific operation: The server organizes the new candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[1347] Step 10:

[1348] The terminal presents the new candidate list to the user.

[1349] Input: The new candidate list data sent from the server.

[1350] Output: The new candidate list displayed in the chat interface.

[1351] Specific operation: The device displays the new candidate list received by the device to the user in the chat interface. For example, "Recommended sweet red wines are Wine D (2600 yen) and Wine E (2900 yen)."

[1352] Step 11:

[1353] The user selects the best wine.

[1354] Input: Final candidates selected by the user through the chat interface.

[1355] Output: The selection results are sent as text data from the terminal to the server.

[1356] Specific operation: The user enters "I will choose wine E," and the device sends the selection result to the server as an API request.

[1357] Step 12:

[1358] The terminal displays a purchase page to the user.

[1359] Input: Purchase page generation instructions from the server.

[1360] Output: Purchase page displayed on the information device.

[1361] Specific operation: The server generates a link to the purchase page and sends it to the device, and the device opens the link and displays the purchase page.

[1362] Step 13:

[1363] The user completes the purchase.

[1364] Input: User's payment information on the purchase page.

[1365] Output: Confirmation that the purchase was successful.

[1366] What happens: The user enters payment information on the purchase page and completes the checkout process.

[1367] Step 14:

[1368] The server processes the payment and arranges shipping.

[1369] Input: Checkout and payment information.

[1370] Output: Confirmation that payment processing has been completed and shipping arrangements have been made.

[1371] Specific operation: The server works with the payment gateway to process the payment and uses the logistics API to arrange for shipping.

[1372] (Application example 1)

[1373] 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."

[1374] Conventional wine selection systems are limited to specific locations, such as the user's home or office. Furthermore, their reliance on smartphones or tablets limits the real-time in-store purchasing experience. Furthermore, while immediate feedback from users is required, there is a lack of efficient means to achieve this. Therefore, there is a need for a system that allows users to efficiently select the optimal wine in-store.

[1375] 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.

[1376] In this invention, the server includes means for receiving conditions input by a user in a chat format using a terminal, a generative AI model for generating candidate wines based on the received conditions, means for displaying a list of candidate wines on the smart glasses, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for displaying the list of re-suggested candidate wines on the smart glasses, and means for generating a procedure page for purchasing the wine selected by the user. This allows the user to select the optimal wine in real time in the store and quickly receive re-suggested wines based on the feedback.

[1377] "User" means an individual or group that uses a service or product.

[1378] A "terminal" is an electronic device used by a user to input conditions, such as a smartphone, tablet, PC, or smart glasses.

[1379] "Chat style" refers to an interface style in which users exchange text and voice in a natural conversational flow.

[1380] "Conditions" refer to specific requests or requirements entered by the user (e.g., budget, preferred flavor, usage scenario, etc.).

[1381] "Receiving" refers to the act of receiving input data or information.

[1382] A "generative AI model" is an artificial intelligence model that is trained to generate optimal candidates based on a user's criteria.

[1383] A "candidate list" is a list of multiple options selected by a generative AI model.

[1384] "Feedback" refers to opinions and additional requests provided by users.

[1385] "Re-proposal" refers to the act of presenting a newly selected candidate based on the feedback received.

[1386] The "procedure page" is a web page where the user inputs and confirms the information required to purchase the selected wine.

[1387] "Smart glasses" are electronic devices in the form of glasses that have a built-in display and camera and have the ability to display information superimposed on the user's field of vision.

[1388] "Display" refers to the act of visually presenting information on the screen of a device or smart glasses.

[1389] "Purchase" refers to the act of obtaining goods or services in exchange for money.

[1390] This invention provides a wine concierge system for efficiently selecting wines that meet the user's needs. Specifically, the user inputs criteria using smart glasses, and a generative AI model generates an optimal candidate list. Based on the feedback, a new candidate list of wines is then displayed, allowing the user to select the optimal wine and proceed with the purchase.

[1391] System Configuration

[1392] The system consists of the following main components:

[1393] 1. User Device:

[1394] A device that allows users to input conditions, including smart glasses. Smart glasses display a chat-style interface and allow voice or touch input. Google Glass is a typical example.

[1395] 2. Server:

[1396] It receives conditions sent from the user's device and passes them to the generative AI model. The server runs in a cloud environment (e.g., an EC2 instance on AWS). Here, it is built as a web application using Flask.

[1397] 3. Generative AI Model:

[1398] It is a generative AI model trained using PyTorch to generate optimal wine candidates based on user input criteria. Specifically, it uses the GPT-3 model.

[1399] Program processing

[1400] The server receives conditional input from the user using a Flask application.

[1401] The received criteria are passed to the GPT-3 model to generate an optimal wine candidate list.

[1402] The server sends the generated wine candidate list to the smart glasses, which then overlays it on the field of view.

[1403] When feedback is received, the conditions are passed to the generative AI model again, and a regenerated candidate list is displayed on the smart glasses.

[1404] Finally, a procedure page for purchasing the wine selected by the user is generated and displayed on the smart glasses.

[1405] Specific examples

[1406] When a user wants to select wine for a house party, the following steps occur:

[1407] 1. Enter the conditions:

[1408] The user voice-inputs into the chat interface of the smart glasses, "I'm looking for wine to use at a home party. My budget is under 4,000 yen, and I'd prefer a white wine."

[1409] 2. Generate wine candidates:

[1410] The server passes the received criteria to the GPT-3 model, which generates an optimal wine candidate list based on the criteria.

[1411] 3. Displaying the results:

[1412] A list of wine candidates is sent from the server to the smart glasses and displayed over the user's field of view. For example, it might look like this: "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[1413] 4. Providing Feedback:

[1414] When the user voices feedback such as "Do you have any sweeter wines?", the server receives this and passes the new criteria to the generative AI model again to generate a revised list of candidates.

[1415] 5. View resubmissions:

[1416] The regenerated candidate list is displayed on the smart glasses. For example, it may look like this: "Candidate 1: Wine C (3,600 yen, sweet)."

[1417] 6. Purchase Procedure:

[1418] The user selects the most suitable wine and inputs the selection result by saying, "I will choose wine C." Based on this, the server generates a checkout page and displays it on the smart glasses.

[1419] Prompt Sentence Examples

[1420] As a concrete example, the prompt sentence "Home party wine budget 4000 yen white wine" is passed to the generative AI model. Based on this prompt sentence, the generative AI model will suggest the most suitable wine candidates.

[1421] The system allows users to use smart glasses to select the perfect wine in real time and quickly receive additional suggestions based on feedback.

[1422] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1423] Step 1:

[1424] The user inputs the requirements by voice into the chat interface of the smart glasses. In this step, the voice is converted into text using a speech recognition API and stored in a database. The specific input is "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable." The output is the textual requirements: "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable."

[1425] Step 2:

[1426] The device sends the textual conditions to the cloud server, which then receives the conditions and prepares them for passing to the generative AI model. The input is the textual conditions, and the output is an HTTP request containing the conditions.

[1427] Step 3:

[1428] The server passes the received conditions to a generative AI model to generate a list of wine candidates. Specifically, it uses a GPT-3 model to generate the best wine candidates based on the conditions. The input is text conditions, and the conditions are converted into prompt sentences as data processing and input into the generative AI model. The output is a list of wine candidates. For example, "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[1429] Step 4:

[1430] The server sends the generated wine candidate list to the smart glasses and displays it over the user's field of view. In this step, data is sent via an HTTP response and the information is displayed on the smart glasses' display. The input is the wine candidate list, and the output is the information overlaid on the smart glasses' display.

[1431] Step 5:

[1432] The user speaks their feedback to the displayed list of candidates, and the feedback is converted into text. For example, they can say, "Do you have any sweeter wines?" and the speech recognition API is used to convert it into text. The input is spoken feedback, and the output is text feedback.

[1433] Step 6:

[1434] The device sends textual feedback to the server, which receives and analyzes it. The conditions are passed to the generative AI model again to generate a new candidate list. The input is the textual feedback, which is converted into a new prompt sentence as data processing and input to the generative AI model. The output is a regenerated list of wine candidates. For example, "Candidate 1: Wine C (3,600 yen, sweet)."

[1435] Step 7:

[1436] The server sends the regenerated wine candidate list to the smart glasses and displays it over the user's field of view. The input is the regenerated wine candidate list, and the output is the information overlaid on the smart glasses' display.

[1437] Step 8:

[1438] The user selects the most suitable wine and inputs the selection by voice or touch. For example, "I will choose wine C." The input is the user's selection, and the output is a text version of the selection.

[1439] Step 9:

[1440] The terminal sends the selection result to the server, and the server generates a checkout page. In this step, a web page containing the information necessary for the checkout process is generated based on the selection result and sent to the smart glasses. The input is the selection result, and the output is the checkout page.

[1441] Step 10:

[1442] The user enters the necessary information on the checkout page displayed on the smart glasses to complete the purchase. Specifically, they enter credit card information and shipping address information. The input is the user's purchase procedure information, and the output is a confirmation message that the purchase has been completed.

[1443] Through these processing steps, users can select the best wine in real time and quickly receive additional suggestions based on feedback.

[1444] 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.

[1445] This invention combines a wine concierge system that utilizes a generative AI model that suggests the most suitable wine based on user input criteria such as budget and preferences when selecting a wine, with an emotion engine that recognizes the user's emotions. Below, the program processing of this system is explained in natural language, with specific examples.

[1446] System Overview

[1447] The system is configured as follows:

[1448] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[1449] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred flavor, usage scenario, etc.).

[1450] 3. The terminal sends the entered conditions to the server as text data.

[1451] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[1452] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[1453] 6. The generative AI model searches a database of wines based on criteria and sentiment to generate the best wine candidates.

[1454] 7. The server organizes the generated candidate list and sends it to the terminal.

[1455] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[1456] 9. The user enters feedback on the presented wine candidates (e.g., "I'd like something sweeter" or "I'm willing to increase my budget")

[1457] 10. The device sends the feedback to the server.

[1458] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[1459] 12. The generative AI model again generates wine suggestions based on criteria and sentiment.

[1460] 13. The server sends the re-proposal candidate list to the terminal.

[1461] 14. The device again presents the chat interface to the user.

[1462] 15. The device will then display a page where the user can select the perfect wine and complete the purchase.

[1463] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[1464] Examples:

[1465] Example 1: Choosing wine for a home party

[1466] 1. The user enters the criteria

[1467] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[1468] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[1469] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[1470] The server receives the conditions and emotion data and passes them to the generative AI model.

[1471] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[1472] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[1473] 3. Users provide feedback

[1474] User: "Do you have any sweeter wines?"

[1475] The device sends this feedback and the user's emotional data to the server.

[1476] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[1477] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[1478] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[1479] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[1480] 5. The user makes the final decision and purchases the wine.

[1481] User: "I'll choose wine E."

[1482] The terminal transmits the selection result to the server.

[1483] The server generates a purchase page and sends it to the terminal.

[1484] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[1485] This allows users to find the perfect wine that reflects their emotions. Through this process, users can significantly reduce the effort and time required to select a wine while also improving their emotional satisfaction.

[1486] The processing flow will be explained below.

[1487] Step 1:

[1488] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[1489] Step 2:

[1490] The user enters the conditions into a chat-style interface. For example, the user might enter, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device then sends this condition to the server as text data.

[1491] Step 3:

[1492] The server receives the conditions. At the same time, the emotion engine analyzes the emotional data from the user's input and generates emotional information such as "joy" or "expectation." This data is then passed to the generative AI model by the server.

[1493] Step 4:

[1494] A generative AI model searches a database of wines based on criteria (budget, red wine) and emotional data (e.g., joy, anticipation). The AI ​​model generates optimal wine candidates and sends them back to the server.

[1495] Step 5:

[1496] The server organizes the wine recommendation list received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, and emotional reasons for the recommendation.

[1497] Step 6:

[1498] The terminal displays the formatted wine candidate list on the chat interface and presents it to the user, who can then confirm the presented candidate list.

[1499] Step 7:

[1500] The user inputs feedback for the presented wine candidates. For example, the user might input, "Do you have any wines that are a little sweeter?" The device then sends this feedback to the server.

[1501] Step 8:

[1502] The server receives the user feedback and analyzes the emotion data again through the emotion engine, which prepares to pass the updated emotion information to the generative AI model.

[1503] Step 9:

[1504] The server passes the feedback and updated emotional data to the generative AI model, which then re-searches for wine candidates based on the new criteria (sweet red wine) and the updated emotional data.

[1505] Step 10:

[1506] The generative AI model generates new wine candidates based on the new criteria and emotional data and sends them back to the server, which receives them and organizes the list of candidates for re-suggestion.

[1507] Step 11:

[1508] The terminal displays the re-suggested wine candidate list in the chat interface and presents it to the user. The re-suggested candidate list includes information based on the new conditions and emotional information.

[1509] Step 12:

[1510] The user selects the best wine. For example, the user inputs, "I will choose wine E." The device sends this selection result to the server.

[1511] Step 13:

[1512] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[1513] Step 14:

[1514] The terminal displays a purchase procedure page to the user using a chat interface, where the user enters the necessary information and completes the purchase procedure.

[1515] Step 15:

[1516] Once the user completes the purchase, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[1517] Step 16:

[1518] The device will display a notification to the user that the purchase is complete and provide information such as the estimated delivery date, allowing the user to select wine efficiently and emotionally in a satisfying way.

[1519] Example 2

[1520] 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."

[1521] Conventional wine selection systems only consider the criteria entered by the user to suggest products, which tends to result in suggestions that ignore the user's feelings. This reduces user satisfaction and makes it difficult to select the appropriate product. In addition, the process of re-suggesting products based on feedback often does not function efficiently, resulting in a suboptimal user experience.

[1522] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing emotions from user input data, means for transferring the analyzed emotion data to the information processing model, and means for analyzing feedback provided by the user and transferring conditions based on the feedback back to the information processing model. This makes it possible to propose products that reflect the user's emotions, improving user satisfaction and realizing an efficient feedback re-proposal process.

[1523] A "user" is a user who uses a terminal to access the service and input conditions and feedback.

[1524] "Terminal" refers to an electronic device used by a user, such as a smartphone, tablet, or PC, and is a device for inputting conditions and feedback through a chat-style interface.

[1525] The "server" is a central system that processes data received from users, performs sentiment analysis, product suggestions and re-suggestions, and sends the results to the terminal.

[1526] The "conditions" refer to the user's wishes and requirements regarding the product he or she wishes to purchase, and specifically include budget, preferred type, purpose, and the like.

[1527] An "information processing model" is an algorithm or system that generates optimal product candidates based on user-entered conditions and sentiment analysis data.

[1528] A "generative AI model" is a type of information processing model, a system that uses AI technology to generate product candidates based on user conditions and emotional data.

[1529] An "emotion engine" is a system that analyzes emotions from user input data and provides the results to a generative AI model.

[1530] "Feedback" refers to additional opinions or requests provided by the user regarding the proposed product candidates, and serves as reference information when the system makes another suggestion.

[1531] A "product candidate list" is a list of candidate products that are judged to meet the conditions from among the group of products proposed by the generative AI model.

[1532] "Re-proposal" refers to the process in which the generative AI model regenerates and proposes a list of product candidates based on user feedback.

[1533] This invention combines an emotion engine that recognizes the user's emotions with a product concierge system that utilizes an information processing model that suggests optimal products based on user input criteria such as budget and preferences when selecting a product. The program processing of this system is explained below, along with specific examples.

[1534] System Overview

[1535] The system is configured as follows:

[1536] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[1537] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred features, usage scenario, etc.).

[1538] 3. The terminal sends the entered conditions to the server as text data.

[1539] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[1540] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[1541] 6. The generative AI model searches the product database based on the conditions and sentiment to generate the best product candidates.

[1542] 7. The server organizes the generated candidate list and sends it to the terminal.

[1543] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[1544] 9. The user enters feedback on the presented product candidates (e.g., "I'd like something sweeter" or "I'd be okay with a higher budget").

[1545] 10. The device sends the feedback to the server.

[1546] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[1547] 12. The generative AI model again generates product candidates based on the conditions and sentiment.

[1548] 13. The server sends the re-proposal candidate list to the terminal.

[1549] 14. The device again presents the chat interface to the user.

[1550] 15. A page will be displayed on the device that allows the user to select the most suitable product and complete the purchase process.

[1551] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[1552] Hardware and software used

[1553] Users use devices such as smartphones, tablets, and PCs.

[1554] The servers are high-performance cloud servers equipped with data processing and storage capabilities.

[1555] The emotion engine uses natural language processing technology to analyze emotions from user input data.

[1556] Generative AI models use machine learning algorithms to suggest the best products based on criteria and sentiment.

[1557] Specific examples

[1558] Example 1: Choosing wine for a home party

[1559] 1. The user enters the criteria

[1560] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[1561] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[1562] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[1563] The server receives the conditions and emotion data and passes them to the generative AI model.

[1564] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[1565] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)" is sent to the terminal.

[1566] 3. Users provide feedback

[1567] User: "Do you have any sweeter wines?"

[1568] The device sends this feedback and the user's emotional data to the server.

[1569] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[1570] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[1571] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[1572] Server: Sends the message "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)" to the terminal.

[1573] 5. The user makes the final decision and purchases the wine.

[1574] User: "I'll choose wine E."

[1575] The terminal transmits the selection result to the server.

[1576] The server generates a purchase page and sends it to the terminal.

[1577] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[1578] This entire process allows users to find the perfect product that reflects their emotions and complete the purchase process smoothly.

[1579] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1580] Step 1:

[1581] The user enters the conditions

[1582] Input: Budget, preferred features, usage scenario, etc.

[1583] Output: Text data containing the condition

[1584] Specific operation: The user opens a chat interface on the device and enters criteria in text format, such as "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device captures this input as text data and immediately sends it to the server.

[1585] Step 2:

[1586] The device sends the user-entered data to the server

[1587] Input: Text data of the condition entered by the user

[1588] Output: Text data sent to the server

[1589] Specific operation: The terminal sends the user's input data to the server via secure communication. The input data is converted into the required format and then passed to the server.

[1590] Step 3:

[1591] The server analyzes the condition and emotion

[1592] Input: Condition data received from the user

[1593] Output: Condition data and emotion data

[1594] Specific operation: The server passes the received text data to the emotion engine, which analyzes the emotions (joy, anticipation, etc.) from the user's input. Condition data and emotion data are generated as the analysis results.

[1595] Step 4:

[1596] The server passes the analysis results to the generative AI model

[1597] Input: Condition data and emotion data

[1598] Output: The data fed into the generative AI model

[1599] Specific operation: The server integrates the emotion data obtained by the emotion engine with the user's condition data and passes it to the generative AI model.

[1600] Step 5:

[1601] Generative AI model generates product candidates

[1602] Input: Condition data and emotion data

[1603] Output: Product candidate list

[1604] Specific operation: The generative AI model searches the database based on the conditions and emotions. For example, if the conditions are "red wine, budget under 3,000 yen," it retrieves appropriate products (e.g., Wine A, Wine B, Wine C) from the database and generates a product candidate list.

[1605] Step 6:

[1606] The server sends the generated candidate list to the device.

[1607] Input: Product candidate list

[1608] Output: Product candidate list sent to the terminal

[1609] Specific operation: The server organizes the generated candidate list, converts it into an easy-to-understand format, and then sends the candidate list to the terminal.

[1610] Step 7:

[1611] The device presents a candidate list to the user.

[1612] Input: Product candidate list sent from the server

[1613] Output: Product candidate list displayed in chat interface

[1614] Specific operation: The device presents the user with a list of options via a chat-style interface, such as "We recommend the following three products: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)."

[1615] Step 8:

[1616] User enters feedback

[1617] Input: Feedback on product candidate list

[1618] Output: Feedback text data

[1619] Specific operation: The user inputs feedback such as "Do you have any wines that are a little sweeter?", and the device captures this feedback as text data and sends it to the server.

[1620] Step 9:

[1621] The device sends feedback to the server

[1622] Input: User feedback text data

[1623] Output: Feedback data sent to the server

[1624] How it works: The device sends feedback to the server over secure communication, and the data is converted into the required format and then passed to the server.

[1625] Step 10:

[1626] The server analyzes the emotions and feedback again.

[1627] Input: Feedback data

[1628] Output: Updated condition data and emotion data

[1629] Specific operation: The server passes the feedback to the emotion engine, which analyzes the emotion again and updates the condition data based on the feedback. For example, if the user's feedback is "sweeter," the condition data is updated to "sweet red wine."

[1630] Step 11:

[1631] The generative AI model generates new product candidates

[1632] Input: Updated condition data and emotion data

[1633] Output: Regenerated product candidate list

[1634] Specific operation: The generative AI model re-searches the database based on new criteria (e.g., sweet red wine) and emotion data to generate new product candidates. If the new list is "Wine D (2,600 yen), Wine E (2,900 yen)," etc., it creates that list.

[1635] Step 12:

[1636] The server sends a list of re-proposals to the device.

[1637] Input: Regenerated product candidate list

[1638] Output: Regeneration candidate list sent to the terminal

[1639] Specific operation: The server organizes the regenerated candidate list, converts it into an easy-to-understand format, and sends it to the terminal.

[1640] Step 13:

[1641] The device presents the candidate list to the user again.

[1642] Input: Regeneration candidate list sent from the server

[1643] Output: Regeneration candidate list displayed in the chat interface

[1644] Specific operation: The device presents the user with a list of regeneration candidates via a chat-style interface, such as "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[1645] Step 14:

[1646] The user decides on the best product and purchases it.

[1647] Input: Final selected product

[1648] Output: Purchase procedure page generated and payment and shipping arrangements completed

[1649] Specific operation: The user enters "I'll choose wine E," and the device sends the selection result to the server. The server generates a purchase procedure page and sends it to the device. The user completes the purchase procedure, and payment and shipping arrangements are completed.

[1650] (Application example 2)

[1651] 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."

[1652] Conventional wine recommendation systems have been unable to select wines that reflect the user's emotions, making it difficult to recommend the best wine for the user. Furthermore, when selecting wine in a brick-and-mortar store, there is a lack of a way for users to efficiently input their budget and preferences and obtain the best wine in real time. The present invention aims to solve these problems and provide a system that can make better wine recommendations that reflect the user's emotions.

[1653] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving conditions entered by a user in a chat format using a terminal; a generative AI model that generates candidate wines based on the received conditions and the user's emotional data; means for transmitting the generated candidate wine list to the terminal; means for receiving feedback provided by the user and again passing the conditions to the generative AI model; means for transmitting the re-proposed candidate wine list to the terminal; means for generating a procedure page for purchasing the wine selected by the user; means for recognizing the user's emotions; means for transferring the emotional data to the generative AI model; and means for using a display device to present the wine selection results to the user. This enables optimal wine recommendations that reflect the user's emotions, enabling efficient wine selection even in physical stores.

[1654] "Means for receiving conditions entered by a user in chat format using a terminal" refers to a mechanism by which a user enters wine selection conditions in chat format using a device such as a smartphone or tablet, and receives that input.

[1655] A "generative AI model that generates candidate wines based on received conditions and user emotional data" is an artificial intelligence model that generates optimal wine candidates based on conditions and emotional data entered by the user.

[1656] The "means for transmitting the generated wine candidate list to the terminal" is a communication function for transmitting the wine candidate list generated by the AI ​​model to the user's device.

[1657] "Means for receiving feedback provided by the user and passing conditions back to the generative AI model" refers to a mechanism for receiving feedback from the user and providing conditions back to the generative AI model based on that feedback.

[1658] The "means for sending a re-proposed wine candidate list to the terminal" is a function for sending a re-generated wine candidate list generated by the AI ​​model based on the updated conditions to the user's device.

[1659] The "means for generating a procedure page for purchasing the wine selected by the user" is a system for generating a web page for purchasing the wine finally selected by the user.

[1660] "Means for recognizing user emotions" refers to technology that analyzes the user's voice, facial expressions, etc. to identify their emotions.

[1661] "Means for transferring emotion data to the generative AI model" refers to a function for transmitting emotion data obtained by emotion recognition to the generative AI model.

[1662] The "means of using a display device to present wine selection results to a user" refers to a technology that uses smart glasses or other display devices to display wine selection results to a user.

[1663] A system for implementing the present invention uses the following hardware and software.

[1664] Hardware:

[1665] Smart devices (smartphones, tablets)

[1666] Smart glasses (e.g., Google Glass)

[1667] server

[1668] software:

[1669] Python

[1670] Flask (web server framework)

[1671] OpenAI API (generative AI model)

[1672] Emotion analysis software (e.g., voice and facial expression recognition libraries)

[1673] System Overview

[1674] The system is configured as follows:

[1675] 1. User input:

[1676] Using a smart device or smart glasses, users can input wine selection criteria (e.g., budget, preferred flavor, intended use, etc.) in a chat format. The device then sends these criteria as data to the server.

[1677] 2. Collecting Emotional Data:

[1678] The smart glasses analyze the user's voice and facial expressions in real time to collect emotional data, which is then sent to the server along with the condition data.

[1679] 3. Sentiment Analysis and Conditional Processing:

[1680] The server analyzes the received conditions and emotion data and uses emotion analysis software to understand the user's emotions.

[1681] 4. Wine selection using generative AI models:

[1682] The server inputs a prompt sentence into the generative AI model based on the emotion data and condition data, and generates wine candidates. For example, the prompt sentence might look like this:

[1683] User sentiment: positive

[1684] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[1685] Choose the three best wines.

[1686] 5. User suggestions for wines:

[1687] The server generates a list of wine candidates, which is then sent to the smart device or smart glasses and displayed to the user. For example, it might show "Wine A (2,500 yen), Wine B (2,800 yen), Wine C (3,000 yen)."

[1688] 6. Feedback and Resubmissions:

[1689] The user inputs feedback on the presented wine candidates (e.g., "A slightly sweeter wine"), and the device sends this feedback and the user's emotional data to the server. The server then performs emotional analysis again, passes new conditions to the generative AI model, and generates a new list of wine candidates, which it then presents to the user.

[1690] 7. Final Decision and Purchase Procedure:

[1691] Once the user selects the wine that best suits them, the device sends the selection results to the server, which then generates a purchase procedure page and displays it to the user. The user completes the purchase procedure on that page, and payment and shipping arrangements are completed.

[1692] Examples:

[1693] Wine selection for a home party:

[1694] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[1695] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[1696] User: "Do you have any sweeter wines?"

[1697] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[1698] User: "I'll choose wine E."

[1699] This system allows users to find the perfect wine that reflects their own feelings, and also streamlines the selection process in physical stores.

[1700] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1701] Step 1:

[1702] The user uses a smart device or smart glasses to input wine selection criteria (e.g., budget, preferred taste, intended use, etc.) in chat format. This generates input data (criteria) from the user.

[1703] Step 2:

[1704] The terminal receives the data entered by the user and sends it to the server. At this time, the data is sent in text format. An example of data sent from the terminal to the server is "Budget: under 3000 yen, red wine for a home party."

[1705] Step 3:

[1706] The smart glasses analyze the user's voice and facial expressions in real time to generate emotional data. The emotional data is analyzed as expressions such as "happiness" or "expectation" using voice and facial recognition software. The generated emotional data is then sent back to the server.

[1707] Step 4:

[1708] The server analyzes the received condition data and emotion data and generates an analysis result using emotion analysis software. Specifically, a voice analysis engine and a facial expression analysis engine are used. Based on this, the user's emotional state is determined to be "positive."

[1709] Step 5:

[1710] The server combines the condition data and emotion data to generate a prompt sentence, which is then input to the generative AI model. An example of this prompt sentence is as follows:

[1711] User sentiment: positive

[1712] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[1713] Choose the three best wines.

[1714] The generative AI model generates a list of optimal wines based on this prompt.

[1715] Step 6:

[1716] The server receives the wine candidate list output from the generative AI model and sends it to the terminal. An example of a wine candidate list that is sent is "Wine A (2500 yen), Wine B (2800 yen), Wine C (3000 yen)."

[1717] Step 7:

[1718] The device receives the wine candidate list and presents it to the user in a chat format. The user then checks the candidate list displayed on the display of the smart glasses or smart device.

[1719] Step 8:

[1720] The user inputs feedback on the wine candidate list, for example, "Do you have any sweeter wines?" This feedback is sent to the server via the terminal.

[1721] Step 9:

[1722] The server analyzes the feedback data and the new emotion data, and again inputs new condition prompts into the generative AI model, which generates a revised wine candidate list. For example, the following prompts are used:

[1723] User sentiment: positive

[1724] Wine selection criteria: A sweet red wine for a home party, within a budget of 3,000 yen

[1725] Choose the three best wines.

[1726] Step 10:

[1727] The server receives the re-proposed wine candidate list and sends it to the terminal. An example of the re-proposed list sent is "Wine D (2600 yen), Wine E (2900 yen)".

[1728] Step 11:

[1729] The terminal again presents the user with a list of wine candidates, and the user selects the most suitable wine. The user then makes a final selection, such as "I will choose wine E."

[1730] Step 12:

[1731] The server generates a procedure page for purchasing the wine selected by the user and sends it to the terminal. The user completes the purchase procedure on the page, and once payment is complete, shipping arrangements are made.

[1732] 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.

[1733] 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.

[1734] 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.

[1735] [Fourth embodiment]

[1736] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1737] 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.

[1738] 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).

[1739] 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.

[1740] 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.

[1741] 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).

[1742] 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.

[1743] 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.

[1744] 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.

[1745] 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.

[1746] 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.

[1747] 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.

[1748] 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."

[1749] This invention is a wine concierge system that utilizes a generative AI model to suggest the most suitable wine based on the user's input criteria such as budget and preferences when selecting a wine. Below, we will explain the program processing of this system in natural language and provide specific examples.

[1750] System Overview

[1751] The system is configured as follows:

[1752] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[1753] 2. The device displays a chat-style interface to the user, and the user enters their requirements (budget, preferred flavor, usage situation, etc.).

[1754] 3. The terminal sends the entered conditions to the server.

[1755] 4. The server receives the conditions and passes the data to the generative AI model.

[1756] 5. The generative AI model generates optimal wine candidates based on the criteria and sends them back to the server.

[1757] 6. The server organizes the generated candidate list and sends it to the terminal.

[1758] 7. The device displays the candidate list on the chat interface and presents it to the user.

[1759] 8. The user enters feedback on the presented options (e.g., "I'd like something a little sweeter" or "I'd be happy to increase the budget").

[1760] 9. The device sends feedback to the server, which again passes the conditions to the generative AI model.

[1761] 10. The generative AI model generates wine candidates again based on the new criteria.

[1762] 11. The server sends the new candidate list to the terminal, which again presents it to the user in the chat interface.

[1763] 12. The device will then display a page where the user can select the perfect wine and complete the purchase.

[1764] 13. Once the user completes the purchase, the server processes the payment and arranges for shipping.

[1765] Examples:

[1766] Example 1: Choosing wine for a home party

[1767] 1. The user enters the criteria

[1768] The user enters, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[1769] The terminal transmits this condition to the server as text data.

[1770] 2. The server receives the conditions and the AI ​​model generates candidates.

[1771] The server receives the conditions and passes them to the generative AI model.

[1772] The generative AI model selects three red wines from a database that meet the criteria.

[1773] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[1774] 3. Users provide feedback

[1775] User: "Do you have any sweeter wines?"

[1776] The terminal sends this feedback to the server.

[1777] 4. The server receives the feedback and the AI ​​model generates candidates again.

[1778] The server receives the feedback and again passes the new conditions to the generative AI model.

[1779] A generative AI model selects a sweet red wine.

[1780] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[1781] 5. The user makes the final decision and purchases the wine.

[1782] User: "I'll choose wine E."

[1783] The terminal transmits the selection result to the server.

[1784] The server generates a purchase page and sends it to the terminal.

[1785] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[1786] This allows users to efficiently and effectively find the wine that suits them, and through this process, users can significantly reduce the effort and time required to select a wine.

[1787] The processing flow will be explained below.

[1788] Step 1:

[1789] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[1790] Step 2:

[1791] The user inputs criteria into a chat-style interface. These criteria include "budget," "preferred flavor," and "usage occasion." The user might input something like, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[1792] Step 3:

[1793] The device sends the entered conditions as text data to the server, which then analyzes the data, converts it into an appropriate format, and prepares it for passing to the generative AI model.

[1794] Step 4:

[1795] The server receives the criteria sent by the user and passes them to the generative AI model, which searches a wine database based on the criteria and generates optimal wine candidates.

[1796] Step 5:

[1797] The generative AI model selects several wine candidates that meet the criteria, for example, three highly rated red wines within a budget of 3,000 yen.

[1798] Step 6:

[1799] The server organizes the candidate list of wines received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, etc.

[1800] Step 7:

[1801] The server sends the formatted wine candidate list to the terminal, which displays the received data on the chat screen and presents wine candidates to the user.

[1802] Step 8:

[1803] The user can input feedback on the presented wine candidates, for example, specifying additional conditions such as "I would prefer something a little sweeter" or "I'm willing to increase my budget a little."

[1804] Step 9:

[1805] The device sends the user's feedback as text data to the server, which then prepares to pass the data to the generative AI model again based on the new conditions.

[1806] Step 10:

[1807] The server receives the feedback from the user, analyzes it, and passes the new criteria to the generative AI model, which then re-searches for wine candidates based on the new criteria.

[1808] Step 11:

[1809] The generative AI model again selects wine candidates that meet the new criteria and sends them back to the server, which then organizes the list of re-suggested wines.

[1810] Step 12:

[1811] The server sends the reformatted list of wine candidates to the terminal, which then displays the received data on the chat screen and re-proposes the wine to the user.

[1812] Step 13:

[1813] The user selects the best wine from the provided candidates and sends the selection result to the server via the terminal.

[1814] Step 14:

[1815] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[1816] Step 15:

[1817] The terminal displays a checkout page to the user in a chat interface, where the user can complete the purchase.

[1818] Step 16:

[1819] Once the user completes the purchase procedure, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[1820] Step 17:

[1821] The device will notify the user of the completion of the purchase and provide information such as the estimated delivery date, allowing the user to select and purchase wine efficiently and effectively.

[1822] Example 1

[1823] 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."

[1824] Conventional wine selection systems make it difficult for users to effectively select wines that match their preferences and requirements, and the selection process takes a lot of time and effort. Furthermore, the process of reflecting user feedback and re-recommending wines is complicated and inefficient. Another problem is that purchasing procedures, payment processing, and shipping arrangements cannot be managed in a single system.

[1825] 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.

[1826] In this invention, the server includes means for receiving conditions input by a user in chat format using an information device, a generative AI model that generates candidate foods and beverages based on the received conditions, means for transmitting the generated candidate food and beverage list to the information device, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for transmitting the re-proposed candidate food and beverage list to the information device, means for generating a procedure page for purchasing the food and beverage selected by the user, means for passing the conditions received from the terminal to the generative AI model in the form of a prompt sentence, means for passing the feedback again to the generative AI model in the form of a prompt sentence when the user provides new feedback, and means for processing payment and arranging for delivery to complete the purchase procedure. This allows users to efficiently and effectively find the food and beverages that suit them and easily complete the purchase procedure.

[1827] An "information device" is a terminal that a user can access (such as a smartphone, tablet, or PC).

[1828] A "generative AI model" is an artificial intelligence algorithm that generates candidate foods and beverages based on received conditions.

[1829] A "prompt format" is a textual description of a question or instruction provided to a generative AI model.

[1830] A "candidate list" is a list of multiple foods and drinks suggested by the generative AI model.

[1831] "Feedback" refers to a reaction, including hopes, requests, corrections, etc., provided by a user to a candidate list.

[1832] "Payment processing" refers to the procedure for completing the payment required to purchase the food and drink selected by the user.

[1833] "Shipping arrangements" are logistics procedures for delivering purchased food and drink to the user.

[1834] "Chat style" refers to an interface between the user and the system that is based on the exchange of text messages.

[1835] "Conditions" are selection criteria entered by the user, such as budget, preferred flavor, and usage scenario.

[1836] The "procedure page" is a web page or app screen for the user to purchase the food and drink selected by the user.

[1837] The present invention is a concierge system that utilizes a generative AI model to suggest optimal food and drink (e.g., wine) based on user input criteria. This system is implemented primarily using information devices, a server, and a generative AI model. Specific embodiments of this system are described below.

[1838] System configuration

[1839] The system is configured as follows:

[1840] 1. Information devices (terminals):

[1841] Devices such as smartphones, tablets, and computers are used.

[1842] These terminals display a chat-style interface for users to enter requirements and provide feedback.

[1843] 2. Server:

[1844] The server implements a generative AI model.

[1845] It receives conditions and feedback entered by the user through the device and passes them on to the generative AI model.

[1846] The generated candidate list is also organized and transmitted to the terminal.

[1847] 3. Generative AI Model:

[1848] Generative AI models (e.g., OpenAI GPT-4 or Transformer-based models) generate optimal food and drink candidates based on criteria and feedback provided by the server.

[1849] The model selects suitable candidates from the database and returns this information to the server.

[1850] System operation example

[1851] As a concrete example, let us consider a scenario of selecting wine for a home party.

[1852] 1. User enters criteria:

[1853] A user enters the following into the chat interface of an information device:

[1854] "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good."

[1855] 2. The server passes the conditions to the generative AI model:

[1856] The server receives this condition and sends it to the generative AI model in the form of a prompt sentence.

[1857] For example: "Please suggest a red wine for a home party within the budget of 3000 yen."

[1858] 3. The generative AI model generates wine candidates:

[1859] Based on the criteria, the generative AI model generates wine candidates such as:

[1860] "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[1861] 4. Users provide feedback:

[1862] The user enters feedback.

[1863] For example: "Do you have any wines that are a little sweeter?"

[1864] 5. The server passes the feedback back to the generative AI model:

[1865] The server then sends the received feedback back to the generative AI model in the form of a prompt sentence.

[1866] For example: "Please suggest a sweeter red wine."

[1867] 6. The generative AI model again generates wine candidates:

[1868] The generative AI model generates new candidates based on the new conditions, such as:

[1869] "The recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[1870] The user selects the most suitable wine from these options, and the terminal generates a purchase page. Once the purchase procedure is complete, the server processes the payment and arranges for delivery.

[1871] In this way, the present invention not only allows the user to efficiently select food and drink, but also allows the user to complete the purchase procedure in an integrated manner, thereby significantly reducing the time and effort required for selecting food and drink.

[1872] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1873] Step 1:

[1874] The user enters the conditions using a terminal.

[1875] Input: A chat interface displayed on an information device (smartphone, tablet, PC). The user types, "I'm looking for wine to use at a home party. My budget is under 3000 yen, and red wine would be good."

[1876] Output: The conditions are sent as text data from the terminal to the server.

[1877] Specific operation: The device receives user input and sends that data to the server as an API request.

[1878] Step 2:

[1879] The server receives the conditions and passes them to the generative AI model in the form of a prompt sentence.

[1880] Input: Conditional text data sent from the terminal.

[1881] Output: The conditions converted into prompt sentence format are passed to the generative AI model.

[1882] Specific operation: The server converts the condition text into a prompt sentence format and inputs it into the generative AI model. It passes it to the model as "Please suggest a red wine for a home party within a budget of 3,000 yen."

[1883] Step 3:

[1884] A generative AI model generates food and drink suggestions based on criteria.

[1885] Input: The prompt text passed by the server.

[1886] Output: A list of food and drink options (wines).

[1887] Specific operation: The generative AI model selects wines from the database that meet the conditions and generates a candidate list. For example, it selects "Wine A (2,500 yen)," "Wine B (2,800 yen)," and "Wine C (3,000 yen)."

[1888] Step 4:

[1889] The server organizes the generated candidate list and transmits it to the terminal.

[1890] Input: A candidate list generated by a generative AI model.

[1891] Output: The curated candidate list is sent to the terminal.

[1892] Specific operation: The server organizes the candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[1893] Step 5:

[1894] The terminal presents the candidate list to the user.

[1895] Input: Candidate list data sent from the server.

[1896] Output: The candidate list displayed in the chat interface.

[1897] Specific operation: The device receives the candidate list and displays it to the user in the chat interface. For example, "We recommend the following three red wines: 1. Wine A (2500 yen) 2. Wine B (2800 yen) 3. Wine C (3000 yen)."

[1898] Step 6:

[1899] The user inputs feedback on the candidate list.

[1900] Input: User-provided feedback through a chat interface. For example, "Do you have any wines that are a little sweeter?"

[1901] Output: Feedback is sent as text data from the device to the server.

[1902] Specific operation: The device receives the user's feedback and sends it to the server as an API request.

[1903] Step 7:

[1904] The server receives the feedback and passes it back to the generative AI model in the form of a prompt sentence.

[1905] Input: Feedback text data sent from the device.

[1906] Output: The feedback converted into a new prompt format is passed to the generative AI model.

[1907] Specific operation: The server converts the feedback text into a prompt sentence format and passes it to the generative AI model as "Please suggest a sweeter red wine."

[1908] Step 8:

[1909] The generative AI model then generates new food and drink suggestions based on the new criteria.

[1910] Input: The new prompt text passed by the server.

[1911] Output: A list of new candidate foods and drinks (wine).

[1912] Specific operation: The generative AI model selects sweet red wines from the database based on the new criteria and generates a candidate list. For example, it selects "Wine D (2,600 yen)" and "Wine E (2,900 yen)."

[1913] Step 9:

[1914] The server organizes a new candidate list and sends it to the terminal again.

[1915] Input: A new candidate list generated by a generative AI model.

[1916] Output: The new, cleaned candidate list is sent to the terminal.

[1917] Specific operation: The server organizes the new candidate list received from the generative AI model, converts it into JSON format, and returns it to the device as an API response.

[1918] Step 10:

[1919] The terminal presents the new candidate list to the user.

[1920] Input: The new candidate list data sent from the server.

[1921] Output: The new candidate list displayed in the chat interface.

[1922] Specific operation: The device displays the new candidate list received by the device to the user in the chat interface. For example, "Recommended sweet red wines are Wine D (2600 yen) and Wine E (2900 yen)."

[1923] Step 11:

[1924] The user selects the best wine.

[1925] Input: Final candidates selected by the user through the chat interface.

[1926] Output: The selection results are sent as text data from the terminal to the server.

[1927] Specific operation: The user enters "I will choose wine E," and the device sends the selection result to the server as an API request.

[1928] Step 12:

[1929] The terminal displays a purchase page to the user.

[1930] Input: Purchase page generation instructions from the server.

[1931] Output: Purchase page displayed on the information device.

[1932] Specific operation: The server generates a link to the purchase page and sends it to the device, and the device opens the link and displays the purchase page.

[1933] Step 13:

[1934] The user completes the purchase.

[1935] Input: User's payment information on the purchase page.

[1936] Output: Confirmation that the purchase was successful.

[1937] What happens: The user enters payment information on the purchase page and completes the checkout process.

[1938] Step 14:

[1939] The server processes the payment and arranges shipping.

[1940] Input: Checkout and payment information.

[1941] Output: Confirmation that payment processing has been completed and shipping arrangements have been made.

[1942] Specific operation: The server works with the payment gateway to process the payment and uses the logistics API to arrange for shipping.

[1943] (Application example 1)

[1944] 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."

[1945] Conventional wine selection systems are limited to specific locations, such as the user's home or office. Furthermore, their reliance on smartphones or tablets limits the real-time in-store purchasing experience. Furthermore, while immediate feedback from users is required, there is a lack of efficient means to achieve this. Therefore, there is a need for a system that allows users to efficiently select the optimal wine in-store.

[1946] 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.

[1947] In this invention, the server includes means for receiving conditions input by a user in a chat format using a terminal, a generative AI model for generating candidate wines based on the received conditions, means for displaying a list of candidate wines on the smart glasses, means for receiving feedback provided by the user and again passing the conditions to the generative AI model, means for displaying the list of re-suggested candidate wines on the smart glasses, and means for generating a procedure page for purchasing the wine selected by the user. This allows the user to select the optimal wine in real time in the store and quickly receive re-suggested wines based on the feedback.

[1948] "User" means an individual or group that uses a service or product.

[1949] A "terminal" is an electronic device used by a user to input conditions, such as a smartphone, tablet, PC, or smart glasses.

[1950] "Chat style" refers to an interface style in which users exchange text and voice in a natural conversational flow.

[1951] "Conditions" refer to specific requests or requirements entered by the user (e.g., budget, preferred flavor, usage scenario, etc.).

[1952] "Receiving" refers to the act of receiving input data or information.

[1953] A "generative AI model" is an artificial intelligence model that is trained to generate optimal candidates based on a user's criteria.

[1954] A "candidate list" is a list of multiple options selected by a generative AI model.

[1955] "Feedback" refers to opinions and additional requests provided by users.

[1956] "Re-proposal" refers to the act of presenting a newly selected candidate based on the feedback received.

[1957] The "procedure page" is a web page where the user inputs and confirms the information required to purchase the selected wine.

[1958] "Smart glasses" are electronic devices in the form of glasses that have a built-in display and camera and have the ability to display information superimposed on the user's field of vision.

[1959] "Display" refers to the act of visually presenting information on the screen of a device or smart glasses.

[1960] "Purchase" refers to the act of obtaining goods or services in exchange for money.

[1961] This invention provides a wine concierge system for efficiently selecting wines that meet the user's needs. Specifically, the user inputs criteria using smart glasses, and a generative AI model generates an optimal candidate list. Based on the feedback, a new candidate list of wines is then displayed, allowing the user to select the optimal wine and proceed with the purchase.

[1962] System Configuration

[1963] The system consists of the following main components:

[1964] 1. User Device:

[1965] A device that allows users to input conditions, including smart glasses. Smart glasses display a chat-style interface and allow voice or touch input. Google Glass is a typical example.

[1966] 2. Server:

[1967] It receives conditions sent from the user's device and passes them to the generative AI model. The server runs in a cloud environment (e.g., an EC2 instance on AWS). Here, it is built as a web application using Flask.

[1968] 3. Generative AI Model:

[1969] It is a generative AI model trained using PyTorch to generate optimal wine candidates based on user input criteria. Specifically, it uses the GPT-3 model.

[1970] Program processing

[1971] The server receives conditional input from the user using a Flask application.

[1972] The received criteria are passed to the GPT-3 model to generate an optimal wine candidate list.

[1973] The server sends the generated wine candidate list to the smart glasses, which then overlays it on the field of view.

[1974] When feedback is received, the conditions are passed to the generative AI model again, and a regenerated candidate list is displayed on the smart glasses.

[1975] Finally, a procedure page for purchasing the wine selected by the user is generated and displayed on the smart glasses.

[1976] Specific examples

[1977] When a user wants to select wine for a house party, the following steps occur:

[1978] 1. Enter the conditions:

[1979] The user voice-inputs into the chat interface of the smart glasses, "I'm looking for wine to use at a home party. My budget is under 4,000 yen, and I'd prefer a white wine."

[1980] 2. Generate wine candidates:

[1981] The server passes the received criteria to the GPT-3 model, which generates an optimal wine candidate list based on the criteria.

[1982] 3. Displaying the results:

[1983] A list of wine candidates is sent from the server to the smart glasses and displayed over the user's field of view. For example, it might look like this: "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[1984] 4. Providing Feedback:

[1985] When the user voices feedback such as "Do you have any sweeter wines?", the server receives this and passes the new criteria to the generative AI model again to generate a revised list of candidates.

[1986] 5. View resubmissions:

[1987] The regenerated candidate list is displayed on the smart glasses. For example, it may look like this: "Candidate 1: Wine C (3,600 yen, sweet)."

[1988] 6. Purchase Procedure:

[1989] The user selects the most suitable wine and inputs the selection result by saying, "I will choose wine C." Based on this, the server generates a checkout page and displays it on the smart glasses.

[1990] Prompt Sentence Examples

[1991] As a concrete example, the prompt sentence "Home party wine budget 4000 yen white wine" is passed to the generative AI model. Based on this prompt sentence, the generative AI model will suggest the most suitable wine candidates.

[1992] The system allows users to use smart glasses to select the perfect wine in real time and quickly receive additional suggestions based on feedback.

[1993] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1994] Step 1:

[1995] The user inputs the requirements by voice into the chat interface of the smart glasses. In this step, the voice is converted into text using a speech recognition API and stored in a database. The specific input is "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable." The output is the textual requirements: "I'm looking for wine to use at a home party. The budget is under 4000 yen, and white wine is preferable."

[1996] Step 2:

[1997] The device sends the textual conditions to the cloud server, which then receives the conditions and prepares them for passing to the generative AI model. The input is the textual conditions, and the output is an HTTP request containing the conditions.

[1998] Step 3:

[1999] The server passes the received conditions to a generative AI model to generate a list of wine candidates. Specifically, it uses a GPT-3 model to generate the best wine candidates based on the conditions. The input is text conditions, and the conditions are converted into prompt sentences as data processing and input into the generative AI model. The output is a list of wine candidates. For example, "Candidate 1: Wine A (3,500 yen), Candidate 2: Wine B (3,800 yen)."

[2000] Step 4:

[2001] The server sends the generated wine candidate list to the smart glasses and displays it over the user's field of view. In this step, data is sent via an HTTP response and the information is displayed on the smart glasses' display. The input is the wine candidate list, and the output is the information overlaid on the smart glasses' display.

[2002] Step 5:

[2003] The user speaks their feedback to the displayed list of candidates, and the feedback is converted into text. For example, they can say, "Do you have any sweeter wines?" and the speech recognition API is used to convert it into text. The input is spoken feedback, and the output is text feedback.

[2004] Step 6:

[2005] The device sends textual feedback to the server, which receives and analyzes it. The conditions are passed to the generative AI model again to generate a new candidate list. The input is the textual feedback, which is converted into a new prompt sentence as data processing and input to the generative AI model. The output is a regenerated list of wine candidates. For example, "Candidate 1: Wine C (3,600 yen, sweet)."

[2006] Step 7:

[2007] The server sends the regenerated wine candidate list to the smart glasses and displays it over the user's field of view. The input is the regenerated wine candidate list, and the output is the information overlaid on the smart glasses' display.

[2008] Step 8:

[2009] The user selects the most suitable wine and inputs the selection by voice or touch. For example, "I will choose wine C." The input is the user's selection, and the output is a text version of the selection.

[2010] Step 9:

[2011] The terminal sends the selection result to the server, and the server generates a checkout page. In this step, a web page containing the information necessary for the checkout process is generated based on the selection result and sent to the smart glasses. The input is the selection result, and the output is the checkout page.

[2012] Step 10:

[2013] The user enters the necessary information on the checkout page displayed on the smart glasses to complete the purchase. Specifically, they enter credit card information and shipping address information. The input is the user's purchase procedure information, and the output is a confirmation message that the purchase has been completed.

[2014] Through these processing steps, users can select the best wine in real time and quickly receive additional suggestions based on feedback.

[2015] 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.

[2016] This invention combines a wine concierge system that utilizes a generative AI model that suggests the most suitable wine based on user input criteria such as budget and preferences when selecting a wine, with an emotion engine that recognizes the user's emotions. Below, the program processing of this system is explained in natural language, with specific examples.

[2017] System Overview

[2018] The system is configured as follows:

[2019] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[2020] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred flavor, usage scenario, etc.).

[2021] 3. The terminal sends the entered conditions to the server as text data.

[2022] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[2023] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[2024] 6. The generative AI model searches a database of wines based on criteria and sentiment to generate the best wine candidates.

[2025] 7. The server organizes the generated candidate list and sends it to the terminal.

[2026] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[2027] 9. The user enters feedback on the presented wine candidates (e.g., "I'd like something sweeter" or "I'm willing to increase my budget")

[2028] 10. The device sends the feedback to the server.

[2029] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[2030] 12. The generative AI model again generates wine suggestions based on criteria and sentiment.

[2031] 13. The server sends the re-proposal candidate list to the terminal.

[2032] 14. The device again presents the chat interface to the user.

[2033] 15. The device will then display a page where the user can select the perfect wine and complete the purchase.

[2034] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[2035] Examples:

[2036] Example 1: Choosing wine for a home party

[2037] 1. The user enters the criteria

[2038] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[2039] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[2040] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[2041] The server receives the conditions and emotion data and passes them to the generative AI model.

[2042] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[2043] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[2044] 3. Users provide feedback

[2045] User: "Do you have any sweeter wines?"

[2046] The device sends this feedback and the user's emotional data to the server.

[2047] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[2048] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[2049] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[2050] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[2051] 5. The user makes the final decision and purchases the wine.

[2052] User: "I'll choose wine E."

[2053] The terminal transmits the selection result to the server.

[2054] The server generates a purchase page and sends it to the terminal.

[2055] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[2056] This allows users to find the perfect wine that reflects their emotions. Through this process, users can significantly reduce the effort and time required to select a wine while also improving their emotional satisfaction.

[2057] The processing flow will be explained below.

[2058] Step 1:

[2059] A user accesses the wine concierge service using a terminal. The terminal displays a chat-style interface to the user, prompting the user to input conditions (budget, preferred taste, usage occasion, etc.).

[2060] Step 2:

[2061] The user enters the conditions into a chat-style interface. For example, the user might enter, "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device then sends this condition to the server as text data.

[2062] Step 3:

[2063] The server receives the conditions. At the same time, the emotion engine analyzes the emotional data from the user's input and generates emotional information such as "joy" or "expectation." This data is then passed to the generative AI model by the server.

[2064] Step 4:

[2065] A generative AI model searches a database of wines based on criteria (budget, red wine) and emotional data (e.g., joy, anticipation). The AI ​​model generates optimal wine candidates and sends them back to the server.

[2066] Step 5:

[2067] The server organizes the wine recommendation list received from the generative AI model and formats it in a user-friendly format, including the wine's name, price, characteristics, and emotional reasons for the recommendation.

[2068] Step 6:

[2069] The terminal displays the formatted wine candidate list on the chat interface and presents it to the user, who can then confirm the presented candidate list.

[2070] Step 7:

[2071] The user inputs feedback for the presented wine candidates. For example, the user might input, "Do you have any wines that are a little sweeter?" The device then sends this feedback to the server.

[2072] Step 8:

[2073] The server receives the user feedback and analyzes the emotion data again through the emotion engine, which prepares to pass the updated emotion information to the generative AI model.

[2074] Step 9:

[2075] The server passes the feedback and updated emotional data to the generative AI model, which then re-searches for wine candidates based on the new criteria (sweet red wine) and the updated emotional data.

[2076] Step 10:

[2077] The generative AI model generates new wine candidates based on the new criteria and emotional data and sends them back to the server, which receives them and organizes the list of candidates for re-suggestion.

[2078] Step 11:

[2079] The terminal displays the re-suggested wine candidate list in the chat interface and presents it to the user. The re-suggested candidate list includes information based on the new conditions and emotional information.

[2080] Step 12:

[2081] The user selects the best wine. For example, the user inputs, "I will choose wine E." The device sends this selection result to the server.

[2082] Step 13:

[2083] The server receives the user's selection, generates a purchase procedure page, and sends the generated page to the terminal.

[2084] Step 14:

[2085] The terminal displays a purchase procedure page to the user using a chat interface, where the user enters the necessary information and completes the purchase procedure.

[2086] Step 15:

[2087] Once the user completes the purchase, the server accesses the payment service and executes the payment process. After the payment is completed, the server arranges for the wine to be shipped.

[2088] Step 16:

[2089] The device will display a notification to the user that the purchase is complete and provide information such as the estimated delivery date, allowing the user to select wine efficiently and emotionally in a satisfying way.

[2090] Example 2

[2091] 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."

[2092] Conventional wine selection systems only consider the criteria entered by the user to suggest products, which tends to result in suggestions that ignore the user's feelings. This reduces user satisfaction and makes it difficult to select the appropriate product. In addition, the process of re-suggesting products based on feedback often does not function efficiently, resulting in a suboptimal user experience.

[2093] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing emotions from user input data, means for transferring the analyzed emotion data to the information processing model, and means for analyzing feedback provided by the user and transferring conditions based on the feedback back to the information processing model. This makes it possible to propose products that reflect the user's emotions, improving user satisfaction and realizing an efficient feedback re-proposal process.

[2094] A "user" is a user who uses a terminal to access the service and input conditions and feedback.

[2095] "Terminal" refers to an electronic device used by a user, such as a smartphone, tablet, or PC, and is a device for inputting conditions and feedback through a chat-style interface.

[2096] The "server" is a central system that processes data received from users, performs sentiment analysis, product suggestions and re-suggestions, and sends the results to the terminal.

[2097] The "conditions" refer to the user's wishes and requirements regarding the product he or she wishes to purchase, and specifically include budget, preferred type, purpose, and the like.

[2098] An "information processing model" is an algorithm or system that generates optimal product candidates based on user-entered conditions and sentiment analysis data.

[2099] A "generative AI model" is a type of information processing model, a system that uses AI technology to generate product candidates based on user conditions and emotional data.

[2100] An "emotion engine" is a system that analyzes emotions from user input data and provides the results to a generative AI model.

[2101] "Feedback" refers to additional opinions or requests provided by the user regarding the proposed product candidates, and serves as reference information when the system makes another suggestion.

[2102] A "product candidate list" is a list of candidate products that are judged to meet the conditions from among the group of products proposed by the generative AI model.

[2103] "Re-proposal" refers to the process in which the generative AI model regenerates and proposes a list of product candidates based on user feedback.

[2104] This invention combines an emotion engine that recognizes the user's emotions with a product concierge system that utilizes an information processing model that suggests optimal products based on user input criteria such as budget and preferences when selecting a product. The program processing of this system is explained below, along with specific examples.

[2105] System Overview

[2106] The system is configured as follows:

[2107] 1. The user accesses the service using a device (smartphone, tablet, PC, etc.).

[2108] 2. The device displays a chat-style interface to the user, prompting them to enter their requirements (budget, preferred features, usage scenario, etc.).

[2109] 3. The terminal sends the entered conditions to the server as text data.

[2110] 4. The server receives the condition and analyzes the user's emotion using the emotion engine.

[2111] 5. The server passes the analyzed emotion data and conditions to the generative AI model.

[2112] 6. The generative AI model searches the product database based on the conditions and sentiment to generate the best product candidates.

[2113] 7. The server organizes the generated candidate list and sends it to the terminal.

[2114] 8. The terminal displays the candidate list on the chat interface and presents it to the user.

[2115] 9. The user enters feedback on the presented product candidates (e.g., "I'd like something sweeter" or "I'd be okay with a higher budget").

[2116] 10. The device sends the feedback to the server.

[2117] 11. The server receives the feedback, analyzes the user's emotions again using the emotion engine, and passes the conditions to the generative AI model.

[2118] 12. The generative AI model again generates product candidates based on the conditions and sentiment.

[2119] 13. The server sends the re-proposal candidate list to the terminal.

[2120] 14. The device again presents the chat interface to the user.

[2121] 15. A page will be displayed on the device that allows the user to select the most suitable product and complete the purchase process.

[2122] 16. Once the user completes the purchase, the server contacts the payment service to process the payment and arrange for shipping.

[2123] Hardware and software used

[2124] Users use devices such as smartphones, tablets, and PCs.

[2125] The servers are high-performance cloud servers equipped with data processing and storage capabilities.

[2126] The emotion engine uses natural language processing technology to analyze emotions from user input data.

[2127] Generative AI models use machine learning algorithms to suggest the best products based on criteria and sentiment.

[2128] Specific examples

[2129] Example 1: Choosing wine for a home party

[2130] 1. The user enters the criteria

[2131] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[2132] The terminal transmits these conditions together with emotion data obtained from the user's input to the server.

[2133] 2. The server receives the conditions and emotion data, and the AI ​​model generates candidates.

[2134] The server receives the conditions and emotion data and passes them to the generative AI model.

[2135] The generative AI model searches a database of wines based on criteria (budget, red wine) and emotions (e.g., joy or anticipation) to select three optimal wine candidates.

[2136] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)" is sent to the terminal.

[2137] 3. Users provide feedback

[2138] User: "Do you have any sweeter wines?"

[2139] The device sends this feedback and the user's emotional data to the server.

[2140] 4. The server receives the feedback and emotion data, and the AI ​​model generates candidates again.

[2141] The server receives the feedback and emotion data and again passes new conditions to the generative AI model.

[2142] The generative AI model re-searches for wine candidates based on the new criteria (sweet red wine) and updated sentiment data.

[2143] Server: Sends the message "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)" to the terminal.

[2144] 5. The user makes the final decision and purchases the wine.

[2145] User: "I'll choose wine E."

[2146] The terminal transmits the selection result to the server.

[2147] The server generates a purchase page and sends it to the terminal.

[2148] The user completes the purchase procedure, and payment and shipping arrangements are completed.

[2149] This entire process allows users to find the perfect product that reflects their emotions and complete the purchase process smoothly.

[2150] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2151] Step 1:

[2152] The user enters the conditions

[2153] Input: Budget, preferred features, usage scenario, etc.

[2154] Output: Text data containing the condition

[2155] Specific operation: The user opens a chat interface on the device and enters criteria in text format, such as "I'm looking for wine to use at a home party. My budget is under 3,000 yen, and red wine would be good." The device captures this input as text data and immediately sends it to the server.

[2156] Step 2:

[2157] The device sends the user-entered data to the server

[2158] Input: Text data of the condition entered by the user

[2159] Output: Text data sent to the server

[2160] Specific operation: The terminal sends the user's input data to the server via secure communication. The input data is converted into the required format and then passed to the server.

[2161] Step 3:

[2162] The server analyzes the condition and emotion

[2163] Input: Condition data received from the user

[2164] Output: Condition data and emotion data

[2165] Specific operation: The server passes the received text data to the emotion engine, which analyzes the emotions (joy, anticipation, etc.) from the user's input. Condition data and emotion data are generated as the analysis results.

[2166] Step 4:

[2167] The server passes the analysis results to the generative AI model

[2168] Input: Condition data and emotion data

[2169] Output: The data fed into the generative AI model

[2170] Specific operation: The server integrates the emotion data obtained by the emotion engine with the user's condition data and passes it to the generative AI model.

[2171] Step 5:

[2172] Generative AI model generates product candidates

[2173] Input: Condition data and emotion data

[2174] Output: Product candidate list

[2175] Specific operation: The generative AI model searches the database based on the conditions and emotions. For example, if the conditions are "red wine, budget under 3,000 yen," it retrieves appropriate products (e.g., Wine A, Wine B, Wine C) from the database and generates a product candidate list.

[2176] Step 6:

[2177] The server sends the generated candidate list to the device.

[2178] Input: Product candidate list

[2179] Output: Product candidate list sent to the terminal

[2180] Specific operation: The server organizes the generated candidate list, converts it into an easy-to-understand format, and then sends the candidate list to the terminal.

[2181] Step 7:

[2182] The device presents a candidate list to the user.

[2183] Input: Product candidate list sent from the server

[2184] Output: Product candidate list displayed in chat interface

[2185] Specific operation: The device presents the user with a list of options via a chat-style interface, such as "We recommend the following three products: 1. Wine A (2,500 yen), 2. Wine B (2,800 yen), 3. Wine C (3,000 yen)."

[2186] Step 8:

[2187] User enters feedback

[2188] Input: Feedback on product candidate list

[2189] Output: Feedback text data

[2190] Specific operation: The user inputs feedback such as "Do you have any wines that are a little sweeter?", and the device captures this feedback as text data and sends it to the server.

[2191] Step 9:

[2192] The device sends feedback to the server

[2193] Input: User feedback text data

[2194] Output: Feedback data sent to the server

[2195] How it works: The device sends feedback to the server over secure communication, and the data is converted into the required format and then passed to the server.

[2196] Step 10:

[2197] The server analyzes the emotions and feedback again.

[2198] Input: Feedback data

[2199] Output: Updated condition data and emotion data

[2200] Specific operation: The server passes the feedback to the emotion engine, which analyzes the emotion again and updates the condition data based on the feedback. For example, if the user's feedback is "sweeter," the condition data is updated to "sweet red wine."

[2201] Step 11:

[2202] The generative AI model generates new product candidates

[2203] Input: Updated condition data and emotion data

[2204] Output: Regenerated product candidate list

[2205] Specific operation: The generative AI model re-searches the database based on new criteria (e.g., sweet red wine) and emotion data to generate new product candidates. If the new list is "Wine D (2,600 yen), Wine E (2,900 yen)," etc., it creates that list.

[2206] Step 12:

[2207] The server sends a list of re-proposals to the device.

[2208] Input: Regenerated product candidate list

[2209] Output: Regeneration candidate list sent to the terminal

[2210] Specific operation: The server organizes the regenerated candidate list, converts it into an easy-to-understand format, and sends it to the terminal.

[2211] Step 13:

[2212] The device presents the candidate list to the user again.

[2213] Input: Regeneration candidate list sent from the server

[2214] Output: Regeneration candidate list displayed in the chat interface

[2215] Specific operation: The device presents the user with a list of regeneration candidates via a chat-style interface, such as "Recommended sweet red wines are Wine D (2,600 yen) and Wine E (2,900 yen)."

[2216] Step 14:

[2217] The user decides on the best product and purchases it.

[2218] Input: Final selected product

[2219] Output: Purchase procedure page generated and payment and shipping arrangements completed

[2220] Specific operation: The user enters "I'll choose wine E," and the device sends the selection result to the server. The server generates a purchase procedure page and sends it to the device. The user completes the purchase procedure, and payment and shipping arrangements are completed.

[2221] (Application example 2)

[2222] 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."

[2223] Conventional wine recommendation systems have been unable to select wines that reflect the user's emotions, making it difficult to recommend the best wine for the user. Furthermore, when selecting wine in a brick-and-mortar store, there is a lack of a way for users to efficiently input their budget and preferences and obtain the best wine in real time. The present invention aims to solve these problems and provide a system that can make better wine recommendations that reflect the user's emotions.

[2224] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving conditions entered by a user in a chat format using a terminal; a generative AI model that generates candidate wines based on the received conditions and the user's emotional data; means for transmitting the generated candidate wine list to the terminal; means for receiving feedback provided by the user and again passing the conditions to the generative AI model; means for transmitting the re-proposed candidate wine list to the terminal; means for generating a procedure page for purchasing the wine selected by the user; means for recognizing the user's emotions; means for transferring the emotional data to the generative AI model; and means for using a display device to present the wine selection results to the user. This enables optimal wine recommendations that reflect the user's emotions, enabling efficient wine selection even in physical stores.

[2225] "Means for receiving conditions entered by a user in chat format using a terminal" refers to a mechanism by which a user enters wine selection conditions in chat format using a device such as a smartphone or tablet, and receives that input.

[2226] A "generative AI model that generates candidate wines based on received conditions and user emotional data" is an artificial intelligence model that generates optimal wine candidates based on conditions and emotional data entered by the user.

[2227] The "means for transmitting the generated wine candidate list to the terminal" is a communication function for transmitting the wine candidate list generated by the AI ​​model to the user's device.

[2228] "Means for receiving feedback provided by the user and passing conditions back to the generative AI model" refers to a mechanism for receiving feedback from the user and providing conditions back to the generative AI model based on that feedback.

[2229] The "means for sending a re-proposed wine candidate list to the terminal" is a function for sending a re-generated wine candidate list generated by the AI ​​model based on the updated conditions to the user's device.

[2230] The "means for generating a procedure page for purchasing the wine selected by the user" is a system for generating a web page for purchasing the wine finally selected by the user.

[2231] "Means for recognizing user emotions" refers to technology that analyzes the user's voice, facial expressions, etc. to identify their emotions.

[2232] "Means for transferring emotion data to the generative AI model" refers to a function for transmitting emotion data obtained by emotion recognition to the generative AI model.

[2233] The "means of using a display device to present wine selection results to a user" refers to a technology that uses smart glasses or other display devices to display wine selection results to a user.

[2234] A system for implementing the present invention uses the following hardware and software.

[2235] Hardware:

[2236] Smart devices (smartphones, tablets)

[2237] Smart glasses (e.g., Google Glass)

[2238] server

[2239] software:

[2240] Python

[2241] Flask (web server framework)

[2242] OpenAI API (generative AI model)

[2243] Emotion analysis software (e.g., voice and facial expression recognition libraries)

[2244] System Overview

[2245] The system is configured as follows:

[2246] 1. User input:

[2247] Using a smart device or smart glasses, users can input wine selection criteria (e.g., budget, preferred flavor, intended use, etc.) in a chat format. The device then sends these criteria as data to the server.

[2248] 2. Collecting Emotional Data:

[2249] The smart glasses analyze the user's voice and facial expressions in real time to collect emotional data, which is then sent to the server along with the condition data.

[2250] 3. Sentiment Analysis and Conditional Processing:

[2251] The server analyzes the received conditions and emotion data and uses emotion analysis software to understand the user's emotions.

[2252] 4. Wine selection using generative AI models:

[2253] The server inputs a prompt sentence into the generative AI model based on the emotion data and condition data, and generates wine candidates. For example, the prompt sentence might look like this:

[2254] User sentiment: positive

[2255] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[2256] Choose the three best wines.

[2257] 5. User suggestions for wines:

[2258] The server generates a list of wine candidates, which is then sent to the smart device or smart glasses and displayed to the user. For example, it might show "Wine A (2,500 yen), Wine B (2,800 yen), Wine C (3,000 yen)."

[2259] 6. Feedback and Resubmissions:

[2260] The user inputs feedback on the presented wine candidates (e.g., "A slightly sweeter wine"), and the device sends this feedback and the user's emotional data to the server. The server then performs emotional analysis again, passes new conditions to the generative AI model, and generates a new list of wine candidates, which it then presents to the user.

[2261] 7. Final Decision and Purchase Procedure:

[2262] Once the user selects the wine that best suits them, the device sends the selection results to the server, which then generates a purchase procedure page and displays it to the user. The user completes the purchase procedure on that page, and payment and shipping arrangements are completed.

[2263] Examples:

[2264] Wine selection for a home party:

[2265] User: "I'm looking for wine to have at a home party. My budget is under 3000 yen, and red wine would be good."

[2266] Server: "We recommend the following three red wines: 1. Wine A (2,500 yen) 2. Wine B (2,800 yen) 3. Wine C (3,000 yen)"

[2267] User: "Do you have any sweeter wines?"

[2268] Server: "We recommend Wine D (2,600 yen) and Wine E (2,900 yen) as sweet red wines."

[2269] User: "I'll choose wine E."

[2270] This system allows users to find the perfect wine that reflects their own feelings, and also streamlines the selection process in physical stores.

[2271] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2272] Step 1:

[2273] The user uses a smart device or smart glasses to input wine selection criteria (e.g., budget, preferred taste, intended use, etc.) in chat format. This generates input data (criteria) from the user.

[2274] Step 2:

[2275] The terminal receives the data entered by the user and sends it to the server. At this time, the data is sent in text format. An example of data sent from the terminal to the server is "Budget: under 3000 yen, red wine for a home party."

[2276] Step 3:

[2277] The smart glasses analyze the user's voice and facial expressions in real time to generate emotional data. The emotional data is analyzed as expressions such as "happiness" or "expectation" using voice and facial recognition software. The generated emotional data is then sent back to the server.

[2278] Step 4:

[2279] The server analyzes the received condition data and emotion data and generates an analysis result using emotion analysis software. Specifically, a voice analysis engine and a facial expression analysis engine are used. Based on this, the user's emotional state is determined to be "positive."

[2280] Step 5:

[2281] The server combines the condition data and emotion data to generate a prompt sentence, which is then input to the generative AI model. An example of this prompt sentence is as follows:

[2282] User sentiment: positive

[2283] Wine selection criteria: A red wine for a home party within a budget of 3,000 yen

[2284] Choose the three best wines.

[2285] The generative AI model generates a list of optimal wines based on this prompt.

[2286] Step 6:

[2287] The server receives the wine candidate list output from the generative AI model and sends it to the terminal. An example of a wine candidate list that is sent is "Wine A (2500 yen), Wine B (2800 yen), Wine C (3000 yen)."

[2288] Step 7:

[2289] The device receives the wine candidate list and presents it to the user in a chat format. The user then checks the candidate list displayed on the display of the smart glasses or smart device.

[2290] Step 8:

[2291] The user inputs feedback on the wine candidate list, for example, "Do you have any sweeter wines?" This feedback is sent to the server via the terminal.

[2292] Step 9:

[2293] The server analyzes the feedback data and the new emotion data, and again inputs new condition prompts into the generative AI model, which generates a revised wine candidate list. For example, the following prompts are used:

[2294] User sentiment: positive

[2295] Wine selection criteria: A sweet red wine for a home party, within a budget of 3,000 yen

[2296] Choose the three best wines.

[2297] Step 10:

[2298] The server receives the re-proposed wine candidate list and sends it to the terminal. An example of the re-proposed list sent is "Wine D (2600 yen), Wine E (2900 yen)".

[2299] Step 11:

[2300] The terminal again presents the user with a list of wine candidates, and the user selects the most suitable wine. The user then makes a final selection, such as "I will choose wine E."

[2301] Step 12:

[2302] The server generates a procedure page for purchasing the wine selected by the user and sends it to the terminal. The user completes the purchase procedure on the page, and once payment is complete, shipping arrangements are made.

[2303] 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.

[2304] 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.

[2305] 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.

[2306] 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.

[2307] 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.

[2308] 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.

[2309] 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).

[2310] 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.

[2311] 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."

[2312] 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.

[2313] The system according to the present disclo...

Claims

1. A means for receiving conditions input by a user in a chat format using a terminal; a generative AI model that generates candidate wines based on the received criteria; a means for transmitting the generated wine candidate list to a terminal; A means of receiving user-provided feedback and passing the conditions back to the generative AI model; a means for transmitting a list of re-suggested wines to a terminal; means for generating a procedure page for purchasing the wine selected by the user; A system including:

2. 10. The system of claim 1, further comprising means for analyzing user-provided feedback and transferring feedback-based conditions to the regenerative AI model.

3. 2. The system according to claim 1, further comprising means for displaying the list of wine candidates and the list of re-suggested wine candidates in a chat format on the user terminal.

Images