System

Abstract

A system is provided.SOLUTION: A system comprising: a means for inputting a user's condition; a server means for receiving the user's condition; a generated AI model means for analyzing big information based on the received condition and generating an optimal candidate commodity list; a server means for formatting the generated candidate commodity list and returning it to the user; and a server means for receiving a feedback from the user, reflecting a condition correction, and calling the generated AI model means again.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] Today's consumers must expend a great deal of time and effort when selecting the best product from a multitude of options, especially for expensive and important products like mattresses. Traditional methods require consumers to visit a store to try out the product or refer to countless reviews online, but these methods are not always reliable ways to find the best product for them. Furthermore, in situations where time is limited, such as when moving or purchasing a new home, quick and effective selection is essential. The challenge is to solve these time-consuming and time-consuming issues consumers encounter when selecting the best mattress. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides a system including a means for inputting user requirement data, a server means for receiving the user requirement data, a generative AI model means for analyzing big data based on the received requirement data and generating an optimal product candidate list, a server means for formatting the generated product candidate list and returning it to the user, and a server means for receiving feedback from the user, reflecting the condition modifications, and again calling the generative AI model means. This system allows users to efficiently find the optimal mattress that meets their requirements. In particular, the chat-style interface allows users to easily and interactively input their requirements and find the optimal product.

[0006] "User" refers to an individual or organization that uses the system, and is the entity that inputs condition data and receives results.

[0007] "Condition data" refers to information entered by the user, and represents the criteria and preferences for selecting a particular product.

[0008] "Input means" refers to an interface or device that allows a user to input condition data into the system.

[0009] "Server" refers to the computing equipment that processes data received from users and interacts with generative AI models.

[0010] "Generative AI model" refers to artificial intelligence technology that analyzes big data and creates an optimal list of product candidates based on the user's criteria.

[0011] "Big data" refers to a large collection of data that is analyzed by a generative AI model as basic data for product selection.

[0012] "Product candidate list" means a list of multiple products selected by the generative AI model based on the user's condition data and provided to the user by the server.

[0013] "Feedback" refers to the act of a user sending additional requests or opinions regarding the proposed products to the server.

[0014] "Condition Modification" means modified condition data that is taken into consideration when the generative AI model regenerates a product candidate list based on user feedback. [Brief explanation of the drawings]

[0015] [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

[0016] 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.

[0017] First, the terms used in the following description will be explained.

[0018] 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).

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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."

[0036] The present invention provides a system that reduces the time and effort required for a user to select an optimal product, particularly a mattress. Specific embodiments of the system are described below.

[0037] Overall system overview

[0038] In this system, users input criteria data via a chat-style interface, and the server uses a generative AI model to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list using the generative AI model again based on that feedback.

[0039] User Action

[0040] Enter the condition:

[0041] Users access a chat-style input screen using their device and enter their desired conditions, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[0042] Server Processing

[0043] Receiving user data:

[0044] The server receives the condition data sent by the user. The received data has the following format:

[0045] json

[0046] {

[0047] "budget": "50000",

[0048] "hardness": "medium",

[0049] "Size": "Single",

[0050] "Brand": "Not specified"

[0051] }

[0052] Invoke the generative AI model:

[0053] The server passes the received condition data to the generative AI model, which analyzes the big data. The generative AI model retrieves relevant mattress data from the database and generates a list of optimal product candidates that match the conditions.

[0054] Generate candidate list:

[0055] Based on the information obtained from the AI ​​model, the server formats the product candidate list and converts it into a format that can be sent back to the user. For example, the following candidate list is generated:

[0056] json

[0057] [

[0058] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0059] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0060] ]

[0061] Sending results to the user:

[0062] The server returns the formatted candidate list to the user's chat screen.

[0063] Feedback Processing

[0064] User feedback:

[0065] The user can review the provided list of candidates and provide feedback on what is closest to their desired fit or on any additional requirements. For example, they could enter, "Mattress A seems good, but let me know if there is anything cheaper."

[0066] Re-invoke the generated AI model:

[0067] The server receives feedback from the user, reflects the revised conditions, and invokes the generative AI model again. This process generates the optimal candidate list based on the new conditions from the database, just as it did last time.

[0068] Generate a new candidate list:

[0069] Based on the revised conditions, the server again reformats the information obtained from the generative AI model and returns a new product candidate list to the user. For example, the following additional candidates are generated:

[0070] json

[0071] [

[0072] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0073] ]

[0074] Final selection and purchase

[0075] Final decision:

[0076] The user finally selects the best product from the presented options and makes a decision, such as "I will purchase Mattress C."

[0077] Purchase procedure:

[0078] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[0079] Specific examples

[0080] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0081] 2. The server receives the condition data, and the generative AI model uses big data to generate the optimal product candidate list.

[0082] 3. The server returns the generated candidate list to the user, presenting "Mattress A" and "Mattress B."

[0083] 4. The user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[0084] 5. The server invokes the generative AI model again based on the feedback and generates a new candidate list.

[0085] 6. The server returns a new candidate, "Mattress C," to the user.

[0086] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[0087] In this way, the system allows users to efficiently find the best mattress.

[0088] The processing flow will be explained below.

[0089] Step 1:

[0090] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[0091] Step 2:

[0092] Server: The server receives the condition data sent by the user. The received data is in the following format:

[0093] json

[0094] {

[0095] "budget": "50000",

[0096] "hardness": "medium",

[0097] "Size": "Single",

[0098] "Brand": "Not specified"

[0099] }

[0100] Step 3:

[0101] Server: The server passes the received condition data to the generative AI model and invokes it. The generative AI model analyzes the condition data using big data and generates an optimal product candidate list.

[0102] Step 4:

[0103] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[0104] Step 5:

[0105] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[0106] json

[0107] [

[0108] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0109] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0110] ]

[0111] Step 6:

[0112] Server: The server returns the formatted product candidate list to the user's chat screen.

[0113] Step 7:

[0114] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[0115] Step 8:

[0116] Server: The server receives feedback from the user, reflects the revised conditions, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[0117] Step 9:

[0118] Generative AI model: The generative AI model analyzes the condition data again and generates a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[0119] Step 10:

[0120] Server: The server formats the new product list and sends it back to the user, for example:

[0121] json

[0122] [

[0123] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0124] ]

[0125] Step 11:

[0126] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[0127] Step 12:

[0128] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[0129] Example 1

[0130] 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."

[0131] In the past, when users selected products such as mattresses, it took a lot of time and effort to find the best product from the vast number of options available. Furthermore, manually searching and comparing product information was cumbersome, making it difficult for users to find the product that best met their requirements.

[0132] 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.

[0133] In this invention, the server includes means for inputting user condition data, data processing means for receiving the user condition data, machine learning model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, data processing means for processing the generated product candidate list and returning it to the user, and data processing means for receiving feedback from the user, reflecting condition modifications, and again calling the machine learning model means, thereby enabling users to efficiently find products that best fit their conditions.

[0134] "User condition data" refers to the desired conditions that a user inputs to select a specific product, and includes data such as budget, hardness, size, brand, etc.

[0135] "Means for inputting user condition data" refers to devices or software that provide an interface for users to input their desired conditions. This includes smartphones, tablets, PC applications, and web browsers.

[0136] "Data processing means for receiving user condition data" refers to a server or network device for receiving and processing the condition data sent by the user.

[0137] "Big data" refers to a huge amount of data, a dataset that can be analyzed to extract valuable information, such as product information, reviews, and pricing data.

[0138] A "machine learning model" is an algorithm or program used to analyze data and generate an optimal list of product candidates based on the results. A specific example is a generative AI model.

[0139] The "generative machine learning model means" refers to software and hardware for receiving user condition data and large-scale data as input and executing a machine learning model to generate an optimal product candidate list.

[0140] "Data processing means" refers to the server and related software for receiving, processing, transmitting and feedback processing data.

[0141] "Product candidate list" refers to a list of product candidates generated based on the user's criteria and presented to the user for consideration.

[0142] "Feedback" refers to the user inputting their opinions and further requirements regarding the provided product candidate list.

[0143] "Condition modification" refers to changing or adding to the original condition data based on user feedback.

[0144] "Invoke" refers to launching a specific function or process (in this case, a machine learning model) to execute it.

[0145] The present invention provides a system for efficiently selecting a desired product, particularly a mattress, by a user. Specific embodiments will be described below.

[0146] Hardware and Software Use

[0147] User condition data input

[0148] The user accesses a chat-style input screen using a web browser or dedicated application on a device (smartphone, tablet, PC, etc.) and enters desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0149] Receiving and processing condition data

[0150] The server receives the condition data sent by the user as an HTTP POST request and parses it in JSON format. The data can be processed using a web server such as NGINX or Apache, or a programming language such as Python or Node.js. The received data is in the following format:

[0151] json

[0152] {

[0153] "budget": "50000",

[0154] "hardness": "medium",

[0155] "Size": "Single",

[0156] "Brand": "Not specified"

[0157] }

[0158] Invoking a generative AI model

[0159] The server invokes a generative AI model (e.g., OpenAI's GPT-3 or Google's BERT) based on the received condition data. The prompt is created as follows:

[0160] "Choose the best mattress for you, with a budget of under ¥50,000, medium firmness, single size, and no particular brand preference."

[0161] Generate product candidate list

[0162] The generative AI model analyzes large-scale data and generates a list of product candidates that match the criteria. The generated list of product candidates is received by the server and formatted to be sent back to the user. Examples of formatting are as follows:

[0163] "Mattress A: Price 48,000 yen, medium firmness, size single, brand X"

[0164] "Mattress B: Price 45,000 yen, medium firmness, size single, brand Y"

[0165] Sending results to the user

[0166] The server returns the formatted product candidate list to the user's chat screen, and the user confirms the provided list.

[0167] Feedback Processing

[0168] The user reviews the provided candidate list and provides feedback on what is closest to their desired outcome or on additional requirements. For example, they might input, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback, reflects the revised requirements, and invokes the generative AI model again. The new prompt text becomes:

[0169] "Choose the best mattress that is medium firm, single size, and doesn't specify the brand, with a budget of 45,000 yen or less, if possible."

[0170] The newly generated product candidate list is formatted in the same way and sent back to the user.

[0171] Specific examples

[0172] When a user inputs "budget under ¥50,000, medium firmness, single size, no particular brand," the server receives the condition data and the generative AI model uses big data to generate an optimal product candidate list. On the other hand, if the user inputs feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," the server readjusts the conditions and sends a new prompt to the generative AI model, generating a new product candidate list.

[0173] In this way, the present invention is a system that efficiently generates optimal product candidates based on condition data, significantly reducing the time and effort required by the user.

[0174] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0175] Step 1:

[0176] The user accesses a chat-style input screen using a device (smartphone, tablet, PC, etc.) and enters the desired conditions. For example, they might enter "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0177] Input: User's desired conditions (budget, hardness, size, brand)

[0178] Output: User's desired conditions data

[0179] Step 2:

[0180] The device sends the user's criteria data to the server, which receives it as an HTTP POST request and parses it into JSON format.

[0181] Input: User's desired conditions data

[0182] Output: Condition data in JSON format

[0183] Step 3:

[0184] The server uses the received condition data to create a prompt for the generative AI model, such as "Please choose the best mattress for me, within a budget of ¥50,000, with medium firmness, single size, and no particular brand preference."

[0185] Input: Condition data in JSON format

[0186] Output: prompt statement

[0187] Step 4:

[0188] The server sends the prompt to a generative AI model (such as OpenAI's GPT-3 or BERT), which analyzes the prompt and generates an optimal product candidate list from large-scale data.

[0189] Input: prompt statement

[0190] Output: Raw product candidate list data

[0191] Step 5:

[0192] The server receives the raw data of the product candidate list returned by the generative AI model and formats it into a format that is easy for users to understand, such as "Mattress A: Price 48,000 yen, Medium firmness, Size single, Brand X."

[0193] Input: Raw data of product candidate list

[0194] Output: A formatted list of product candidates

[0195] Step 6:

[0196] The server returns the formatted product candidate list to the user's chat screen, which the user can confirm.

[0197] Input: A formatted list of product candidates

[0198] Output: A list of product candidates presented to the user

[0199] Step 7:

[0200] The user enters feedback on the presented product candidate list. For example, "Mattress A seems good, but please let me know if there is anything cheaper."

[0201] Input: User feedback

[0202] Output: Feedback data

[0203] Step 8:

[0204] The device sends the user's feedback data to the server. The server receives this data and modifies the condition data. It then creates a new prompt and sends it back to the generative AI model. For example, it generates a prompt such as, "Please choose the best mattress for me, preferably within 45,000 yen, with medium firmness, single size, and no particular brand preference."

[0205] Input: Feedback data

[0206] Output: Modified prompt statement

[0207] Step 9:

[0208] The generative AI model generates a new product candidate list based on the new prompt. The server receives this and formats it again. The formatted list is then sent to the user again.

[0209] Input: modified prompt text

[0210] Output: Regenerated product candidate list

[0211] Step 10:

[0212] Finally, the user selects the most suitable product from the presented product candidates and notifies the system of their intention to purchase the selected product. The server provides the user with information regarding the purchase procedure.

[0213] Input: Final selected product data

[0214] Output: Purchase procedure guide information

[0215] (Application example 1)

[0216] 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."

[0217] Current online shopping sites have the problem that it is difficult for users to efficiently find the products they want. In particular, because they do not provide prompt responses to input of conditions or feedback, users must spend a great deal of time and effort to find the optimal product. The purpose of this invention is to solve these problems and provide a system that allows users to efficiently select the optimal product.

[0218] 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.

[0219] In this invention, the server includes means for inputting user condition data, means for receiving the user condition data, a generation AI model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, means for formatting the generated product candidate list and returning it to the user, means for receiving feedback from the user and reflecting the condition modifications to call the generation AI model means again, a user interface means using a smartphone, and means for presenting the generated product candidate list to the user and providing feedback on additional conditions, thereby enabling the user to quickly and efficiently select optimal products.

[0220] "User condition data" refers to the desired conditions entered by the user when selecting a product, such as budget, size, hardness, brand, and other specific requirements.

[0221] "Input means" refers to the interface or device through which the user inputs condition data, such as a smartphone, PC, or tablet.

[0222] "Server means" refers to a computer system for receiving, storing, and analyzing data sent by users, including databases and cloud servers.

[0223] "Generative AI model means" refers to an artificial intelligence model used to analyze received data and generate optimal product candidates, including, for example, a machine learning model or a natural language processing model.

[0224] "Product candidate list" refers to data that lists multiple candidate products generated by a generative AI model based on user condition data.

[0225] "Formatting" refers to the process of converting the generated product candidate list into a format that is easy for users to view and understand.

[0226] "Receiving feedback" refers to the process of receiving additional conditions, opinions, etc. from users via input devices.

[0227] "Condition Modification" refers to the process of changing or modifying the original condition data based on feedback received from users.

[0228] "Invoking the generative AI model means again" refers to the process of generating a new product candidate list using the generative AI model again based on the modified condition data.

[0229] "Smartphone-based user interface means" refers to an interface that allows a user to interact with the system via a smartphone, input conditions, and send feedback.

[0230] "Means for providing feedback on additional conditions" refers to a process or interface that allows users to input additional requirements or requested changes and send them to the server.

[0231] System Overview

[0232] The present invention provides a system that allows users to select optimal products, particularly through a smartphone-based user interface, enabling them to find products efficiently and quickly. This system inputs user criteria data and generates an optimal product candidate list using a generative AI model based on that data. The system then presents the generated product candidate list to the user, receives feedback, and uses the generative AI model again to provide the optimal products that meet the user's needs.

[0233] Hardware and Software

[0234] The user's smartphone acts as the input interface, and the server receives the data, runs the generative AI model, generates a list of product candidates, and processes the feedback. For generative AI models, OpenAI GPT-4 and Google BERT are used as examples.

[0235] Program processing

[0236] 1. Enter your user data:

[0237] Users use their smartphones to access the chatbot screen on a dedicated application.

[0238] Users can input the necessary criteria for product selection via text or voice, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[0239] 2. Sending and receiving condition data:

[0240] The smartphone application converts the condition data entered by the user into JSON format and sends it to the server.

[0241] The server analyzes the received condition data and inputs it into the generative AI model.

[0242] 3. Run the generative AI model:

[0243] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[0244] The generative AI model searches and analyzes relevant product information from an internal database to extract the most suitable candidates.

[0245] 4. Generate and return product candidate list:

[0246] The server formats the generated product candidate list and converts it into a format that is easy for the user to understand.

[0247] The formatted product candidate list is sent back to the smartphone application and displayed on the user's screen.

[0248] Specific examples

[0249] If a user types "I'm looking for a medium-firm mattress under 50,000 yen" into a smartphone app, the system will operate as follows:

[0250] 1. A user types into a smartphone app, "I'm looking for a medium-firm mattress under 50,000 yen."

[0251] 2. This text data is converted to JSON format and sent to the server.

[0252] 3. The server inputs the data into a generative AI model (e.g., OpenAI GPT-4 or Google BERT) to generate a list of product candidates that match the criteria.

[0253] 4. Format the generated product candidate list and return it to the user in the following format:

[0254] "Mattress A, Price: 48,000 yen, Firmness: Medium, Size: Single, Brand: Brand X"

[0255] "Mattress B, Price: 45,000 yen, Firmness: Medium, Size: Single, Brand: Brand Y"

[0256] Prompt Sentence Examples

[0257] An example of a prompt sentence to input to a generative AI model is as follows.

[0258] "I'm looking for a mattress. My budget is under ¥50,000, I want a medium-firm mattress, and I don't have a specific brand in mind. Can you recommend the best product for me?"

[0259] In this way, it is possible to realize a system that allows users to quickly and efficiently select the most suitable product.

[0260] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0261] Step 1:

[0262] The user accesses the chatbot screen of the dedicated application using a smartphone. The user inputs the necessary product selection criteria by text or voice. For example, "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0263] Input: The condition data entered by the user.

[0264] Output: Condition data collected on the user's device.

[0265] Step 2:

[0266] The terminal converts the condition data entered by the user into JSON format and sends it to the server.

[0267] Input: Condition data.

[0268] Data processing: Conversion to JSON format.

[0269] Output: Criteria data converted to JSON format.

[0270] Step 3:

[0271] The server analyzes the received JSON formatted condition data and inputs it into the generative AI model.

[0272] Input: Criteria data in JSON format.

[0273] Data analysis: Parsing JSON data and applying it to generative AI models.

[0274] Output: Analysis results and inputs to generative AI models.

[0275] Step 4:

[0276] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[0277] Input: Parsed condition data.

[0278] Data computation: Generative AI models generate product candidate lists.

[0279] Output: Product candidate list.

[0280] Step 5:

[0281] The generated product candidate list is formatted on the server and converted into a format that is easy for users to understand.

[0282] Input: Product picklist.

[0283] Data processing: Formatting the product candidate list.

[0284] Output: A formatted list of product suggestions.

[0285] Step 6:

[0286] The server returns the formatted product candidate list to the user's smartphone application, which displays it on the user's screen.

[0287] Input: A formatted list of product suggestions.

[0288] Output: Product candidate list displayed on user's device.

[0289] Step 7:

[0290] The user checks the product candidate list, sets additional requirements and feedback, and re-enters the list. An example of feedback is, "Mattress A is good, but I want something a little cheaper."

[0291] Input: User feedback.

[0292] Output: Feedback data with added and modified conditions.

[0293] Step 8:

[0294] The server again receives feedback from the user and generates new JSON data that reflects the modified conditions.

[0295] Input: Feedback data.

[0296] Data processing: Generate JSON data that reflects the condition modifications.

[0297] Output: The modified JSON data.

[0298] Step 9:

[0299] Based on the revised condition data, the server calls the generative AI model again and generates a new list of product candidates.

[0300] Input: The modified condition data.

[0301] Data computation: Generative AI models generate new product candidate lists.

[0302] Output: A new list of product candidates.

[0303] Step 10:

[0304] The server then formats the regenerated product candidate list, converts it into a format that is easy for the user to understand, and returns it to the user's smartphone application.

[0305] Input: A new product shortlist.

[0306] Data processing: Formatting a new product candidate list.

[0307] Output: The formatted new product candidate list displayed on the user's device.

[0308] This process creates a system that allows users to efficiently select the most suitable product.

[0309] 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.

[0310] The present invention provides a system that reduces the time and effort required for users to select optimal products, especially mattresses. A key feature of the system is the incorporation of an emotion engine that recognizes the user's emotions and selects products that reflect their emotional state, enabling more personalized recommendations. Specific embodiments of the system are described below.

[0311] Overall system overview

[0312] In this system, users input criteria data via a chat-style interface, and the server then uses a generative AI model and an emotion engine to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[0313] User Action

[0314] Enter the condition:

[0315] Users access a chat-style input screen using their device and input their desired conditions (budget, firmness, size, brand, etc.) For example, they could input "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0316] Server Processing

[0317] Receiving user data:

[0318] The server receives the condition data sent by the user. The received data has the following format:

[0319] json

[0320] {

[0321] "budget": "50000",

[0322] "hardness": "medium",

[0323] "Size": "Single",

[0324] "Brand": "Not specified"

[0325] }

[0326] Calling the Emotion Engine:

[0327] Based on the received condition data, the server calls the emotion engine and analyzes the user's emotional state. The emotion engine recognizes emotions through chat context, user input, facial recognition, and voice analysis.

[0328] Processing generative AI models

[0329] Invoke the generative AI model:

[0330] The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[0331] Generate product picklist:

[0332] The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria and emotional state. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[0333] Server Processing

[0334] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[0335] json

[0336] [

[0337] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0338] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0339] ]

[0340] Sending results to the user:

[0341] The server returns the formatted product candidate list to the user's chat screen.

[0342] Feedback Processing

[0343] User feedback:

[0344] The user can review the provided list of product candidates and provide feedback on what is closest to their desired product or on any additional requirements. For example, they can enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[0345] Re-invoke the generated AI model:

[0346] The server receives feedback from the user, reflects the modified conditions and emotional state, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[0347] Regenerate product pick list

[0348] Generative AI models:

[0349] The generative AI model then analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[0350] server:

[0351] The server then formats the new product list and sends it back to the user, for example:

[0352] json

[0353] [

[0354] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0355] ]

[0356] Final selection and purchase

[0357] Final decision:

[0358] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[0359] Purchase procedure:

[0360] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details of the process.

[0361] Specific examples

[0362] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0363] 2. The server receives the condition data, and the emotion engine recognizes the user's emotional state. At this time, it determines that the user is feeling stressed.

[0364] 3. The generative AI model uses big data to analyze the conditions and emotional state and suggests mattresses with relaxing effects.

[0365] 4. The server returns the generated candidate lists "Mattress A" and "Mattress B" to the user.

[0366] 5. A user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[0367] 6. The server passes the feedback and emotional state back to the generative AI model and suggests a new candidate, "Mattress C."

[0368] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[0369] The system allows users to find the perfect mattress in an efficient and emotionally sensitive manner.

[0370] The processing flow will be explained below.

[0371] Step 1:

[0372] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[0373] Step 2:

[0374] Server: The server receives the condition data sent by the user. The received data is in the following format:

[0375] json

[0376] {

[0377] "budget": "50000",

[0378] "hardness": "medium",

[0379] "Size": "Single",

[0380] "Brand": "Not specified"

[0381] }

[0382] Step 3:

[0383] Server: Based on the received condition data, the server calls the emotion engine to analyze the user's emotional state. The emotion engine recognizes the user's emotions using the user's input and other emotional indicators (e.g., context, facial recognition, voice analysis, etc.).

[0384] Step 4:

[0385] Emotion engine: The emotion engine recognizes the user's emotional state, for example, determining that the user is feeling stressed. The recognized emotional state is returned to the server.

[0386] Step 5:

[0387] Server: The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[0388] Step 6:

[0389] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the user's criteria and emotional state. This process involves running database queries, filtering, and finally ranking the products by scoring.

[0390] Step 7:

[0391] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be returned to the user. For example, the following candidate list is generated:

[0392] json

[0393] [

[0394] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0395] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0396] ]

[0397] Step 8:

[0398] Server: The server returns the formatted product candidate list to the user's chat screen.

[0399] Step 9:

[0400] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[0401] Step 10:

[0402] Server: The server receives feedback from the user and passes it back to the generative AI model along with condition modifications and emotional states.

[0403] Step 11:

[0404] Generative AI model: The generative AI model analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[0405] Step 12:

[0406] Server: The server formats the new product list and sends it back to the user, for example:

[0407] json

[0408] [

[0409] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0410] ]

[0411] Step 13:

[0412] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[0413] Step 14:

[0414] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[0415] Example 2

[0416] 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."

[0417] Conventional product recommendation systems only suggest products based on the user's criteria, without taking into account the user's emotional state. This creates a problem in that they provide products without understanding the user's feelings when searching for a product. As a result, the suggested products often do not match the user's actual needs, making it difficult to increase satisfaction.

[0418] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an input device means for inputting user condition data, a server device means for receiving the user condition data, a generation AI model means for generating an optimal product candidate list based on the received condition data and the user's emotional state recognized by the emotion analysis means, a server device means for shaping the generated product candidate list and returning it to the user, and a server device means for receiving feedback from the user, reflecting the condition modifications and emotional state, and again calling the generation AI model means. This enables personalized product recommendations that take the user's emotional state into consideration.

[0419] "User" means a person who uses the System to select and purchase a Product.

[0420] "Condition data" is information about desired conditions (budget, hardness, size, brand, etc.) that a user enters into the system.

[0421] The "input device means" is a device (such as a personal computer or smartphone) that the user uses to input condition data.

[0422] "Server device means" refers to a server device for receiving and processing data from a user.

[0423] "Emotion analysis means" refers to technology that recognizes the user's emotional state through user input, facial recognition, voice analysis, etc.

[0424] The "generative AI model means" is an artificial intelligence model that analyzes big data based on the user's condition data and emotional state, and generates an optimal product candidate list.

[0425] A "product candidate list" is a plurality of candidate products selected by a generative AI model means and suggested to a user.

[0426] "Feedback" refers to the act of a user providing more desirable or additional conditions for a product candidate list.

[0427] "Condition modification" is the process of changing and updating the initial conditions based on user feedback.

[0428] "Invoking the generating AI model means again" is a procedure for causing the generating AI model means to generate a product candidate list again based on the condition modification and emotional state.

[0429] This invention relates to a system that reduces the time and effort required for users to select the optimal product, especially a mattress. This system is capable of providing more personalized recommendations by incorporating an emotion engine that recognizes the user's emotions and selects products that reflect the user's emotional state.

[0430] Overall system overview

[0431] In this system, users input criteria data through a chat-style interface, and the server generates and provides an optimal product candidate list using a generative AI model and emotion engine based on that data. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[0432] Specific Embodiments

[0433] User input operations

[0434] The user accesses the chat screen of the system using a terminal and inputs the desired data. For example, the user can input specific desired data such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference." This input information is input in text format as follows:

[0435] Budget: 50,000 yen or less, medium hardness, single size, no particular brand preference.

[0436] Server reception and analysis

[0437] The server receives the condition data sent by the user and analyzes it. The data is sent to the server in JSON format, and then the emotion engine is called to analyze the user's emotional state. The emotion engine uses the device's camera and microphone to perform facial recognition and voice analysis, and obtains an analysis result such as "the user is feeling stressed."

[0438] Invoking a generative AI model

[0439] The server passes the analysis results and condition data to the AI ​​model to generate the optimal product candidate list. The input prompt is set as follows:

[0440] Budget: ¥50,000 or less, medium hardness, single size, no particular brand. Users are feeling stressed.

[0441] The generative AI model analyzes large amounts of data and selects the best product candidates based on criteria and emotional state.

[0442] Generate and provide product candidate lists

[0443] The product candidate list generated by the generative AI model is formatted on the server and provided to the user. Specifically, the server replies to the user's chat screen with a message like the following:

[0444] The recommended mattresses are as follows:

[0445] 1. Mattress A - Price: 48,000 yen - Firmness: Medium - Size: Single - Brand X

[0446] 2. Mattress B - Price: 45,000 yen - Firmness: Medium - Size: Single - Brand Y

[0447] Receiving feedback and reanalyzing

[0448] The user reviews the provided product candidate list and provides feedback with more detailed requirements. For example, they may send feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback and calls the generative AI model again to begin the process of generating a new product candidate list.

[0449] Providing a regenerated candidate list

[0450] The server again formats the new product candidate list obtained from the generative AI model and provides it to the user in the following format:

[0451] This is a new proposal.

[0452] 1. Mattress C - Price: 42,000 yen - Firmness: Medium - Size: Single - Brand: Z

[0453] Final selection and checkout

[0454] The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C." The server receives the selection and guides the user to detailed information about the purchase procedure and a link to the purchase page.

[0455] This system allows users to efficiently select the most suitable product while taking into account their own emotional state.

[0456] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0457] Step 1: User Input

[0458] Users access the chat screen of the system using a terminal and input their desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0459] Input: User's desired conditions (budget, hardness, size, brand, etc.)

[0460] Output: The user's desired conditions are sent from the terminal in text format.

[0461] Step 2: Server Receive

[0462] The server receives the condition data sent by the user and prepares to analyze the received data.

[0463] Input: The user's desired conditions are sent to the server in text format.

[0464] Output: The criteria data is converted to JSON format and passed to the next processing step.

[0465] Step 3: Sentiment Analysis

[0466] The server calls the emotion engine based on the condition data and analyzes the user's emotional state. The emotion engine uses the device's camera and microphone to recognize the user's face and analyze their voice.

[0467] Input: Condition data (JSON format) and user input

[0468] Output: Sentiment analysis result (e.g., user is stressed)

[0469] Step 4: Call the generative AI model

[0470] The server passes the analyzed emotional state and condition data to a generative AI model, which then generates an optimal product candidate list. The generative AI model performs analysis using big data.

[0471] Input: Condition data and emotion analysis results

[0472] Output: Product candidate list generated by the generative AI model

[0473] Step 5: Product candidate list formatting

[0474] The server formats the product candidate list obtained from the generative AI model and prepares it for return to the user.

[0475] Input: Product candidate list generated by the generative AI model

[0476] Output: A formatted list of product candidates

[0477] Step 6: Reply to the user

[0478] The server sends the formatted product candidate list to the user's chat screen.

[0479] Input: A formatted list of product candidates

[0480] Output: Product candidate list displayed on the user's chat screen

[0481] Step 7: User feedback

[0482] Users can review the provided product shortlist and enter additional preferences and feedback, such as "Mattress A looks good, but let me know if there's anything cheaper."

[0483] Input: User additional requirements and feedback

[0484] Output: User feedback sent from the terminal in text format

[0485] Step 8: Server reception and reparsing

[0486] The server receives feedback sent by the user and invokes a regenerative AI model that reflects condition modifications and emotional states.

[0487] Input: User feedback (text format)

[0488] Output: Reanalyzed condition data and sentiment analysis results

[0489] Step 9: Regenerate AI model call

[0490] The server then calls the generative AI model again based on the reanalyzed condition data and emotional state, and generates a new list of product candidates.

[0491] Input: Reanalyzed condition data and emotional state

[0492] Output: A new product candidate list generated by the generative AI model

[0493] Step 10: Provide a regenerated candidate list

[0494] The server formats the regenerated product candidate list and provides it to the user's chat screen again.

[0495] Input: A new product candidate list generated by the regenerative AI model

[0496] Output: New, re-formatted product candidate list

[0497] Step 11: Final user selection

[0498] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[0499] Input: User's final selection

[0500] Output: The final selected product information is sent to the server.

[0501] Step 12: Purchase procedure guide

[0502] The server receives the final selected product information and provides the user with detailed purchase procedure information and a link to the purchase page.

[0503] Input: User's last selected product information

[0504] Output: A purchase procedure message will be displayed on the user's chat screen.

[0505] By combining the above steps, users can efficiently select the most suitable mattress and proceed with the purchase process while taking into account their own emotional state.

[0506] (Application example 2)

[0507] 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."

[0508] Conventional product selection systems suggest products based on user condition data, but because they do not take the user's emotional state into account, they have the problem of not being able to make suggestions that are optimal for the user's psychological state. In addition, there is no way to display product location information within the store, making it difficult for users to search for products efficiently. This reduces users' motivation to purchase, and there are issues such as the time and effort required to select the most suitable product.

[0509] 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: an input means for inputting user condition data; a communication means for receiving the user condition data; a generation AI model means for analyzing the received condition data and emotion analysis data to generate an optimal product candidate list; a communication means for formatting the generated product candidate list and returning it to the user; a communication means for reflecting condition modifications based on user feedback and emotion analysis data and again calling the generation AI model means; an emotion recognition means for recognizing the user's emotional state in real time; and a visualization means for displaying product locations within a physical store for the user. This enables personalized product suggestions based on the user's emotional state, allowing the user to efficiently search for products within the store.

[0510] "User condition data" refers to the desired conditions (budget, hardness, size, brand, etc.) entered by the user.

[0511] "Input means" refers to a device or interface that allows a user to input condition data.

[0512] "Communication means" refers to a network interface for sending and receiving data.

[0513] "Emotion analysis data" is data that represents the user's emotional state after analyzing it.

[0514] "Generative AI model means" refers to an artificial intelligence model for generating an optimal product candidate list using received data.

[0515] "Emotion recognition means" refers to technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state in real time.

[0516] "Visualization means" refers to displays and AR technologies that allow users to visually view information.

[0517] "Product Candidate List" means a list of optimal products generated by the generative AI model means and based on the user's criteria and emotional state.

[0518] "Feedback" refers to any additional preferences or comments provided by the User.

[0519] "Physical store" refers to an actual store space.

[0520] To implement this invention, a system including the following components is required. This system is designed to suggest optimal products based on a user's condition data and emotion data. Specifically, it uses a user input means, a communication means, an emotion recognition means, a generation AI model means, and a visualization means.

[0521] First, the user puts on the smart glasses and inputs their requirements by voice or touch. This includes information such as budget, hardness, size, and brand. For example, "budget is under ¥50,000, hardness is medium, size is single, and brand is not specified."

[0522] The input condition data is sent to the server via a communication means. The server receives the data and analyzes the user's emotional state via an emotion recognition means. Emotion recognition uses the smart glasses' camera and microphone to analyze the user's facial expressions and tone of voice. For example, it can determine whether the user is feeling stressed.

[0523] The server then passes the received condition data and emotion data to the generative AI model means, which analyzes this information using big data. The generative AI model generates an optimal product candidate list based on the condition and emotion state. The generated list is sent back to the smart glasses via the communication means.

[0524] The smart glasses use a visualization tool to present the generated product candidate list to the user, along with the product location information, allowing the user to efficiently search for products in the store.

[0525] For example, if "Mattress A" is included in the generated candidate list, its location information will be displayed on the display. The user can walk around the store while checking this information and check the products directly. If the user gives feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," this feedback will also be sent to the server via the communication means.

[0526] The server then calls the generative AI model again based on the received feedback and emotional data, updating the product candidate list. In this way, the user's feedback and emotional state are reflected in real time, allowing for smooth selection of the most suitable product.

[0527] As a concrete example, the following prompt sentence is input to the generative AI model:

[0528] "List products (mattresses) that match the following criteria: Conditions: { "Budget": "50000", "Firmness": "Medium", "Size": "Single", "Brand": "Not Specified"} User emotional state: "Feeling stressed""

[0529] The system allows users to find the perfect mattress in an efficient and emotionally sensitive way.

[0530] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0531] Step 1:

[0532] The user puts on the smart glasses and inputs the desired conditions by voice or touch operation. The input conditions include budget, hardness, size, brand, etc. Data such as "budget within 50,000 yen, hardness medium, size single, no particular brand" is input. The input data is temporarily stored in the smart glasses' internal memory.

[0533] Step 2:

[0534] The user's input data is sent to the server via a communication method, either via the internet or the store's Wi-Fi. The server processes the received condition data as data for analysis. At this time, the received data is converted to JSON format and saved in the server's data storage.

[0535] Step 3:

[0536] The server then calls the emotion recognition unit based on the received data. It uses the smart glasses' camera and microphone to collect the user's facial expressions and tone of voice, analyzes them, and recognizes their emotional state. The analyzed emotional state is then classified into categories such as "feeling stressed" and output as data.

[0537] Step 4:

[0538] The server passes the condition data and the emotion data obtained by emotion recognition to the generative AI model means. The generative AI model receives these data as input and performs analysis using big data. The analysis process includes running database queries, filtering, and scoring. The generative AI model outputs a list of optimal product candidates, which is sent back to the server in JSON format.

[0539] Step 5:

[0540] The server then formats the received product candidate list and converts it into a data format for sending back to the user. This conversion process includes formatting data such as product name, price, and location information. The formatted data is then sent back to the smart glasses via a communication means.

[0541] Step 6:

[0542] The smart glasses display the received product candidate list to the user using a visualization means. The smart glasses display shows the names, prices, and location information of the candidate products in the store. The user confirms the displayed information and moves around the store to actually check the products.

[0543] Step 7:

[0544] After checking the products, the user can provide feedback by voice input or touch operation on the smart glasses. For example, they can input feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." This feedback is sent to the server via a communication means.

[0545] Step 8:

[0546] The server receives the feedback and updates the condition data and emotion data. Based on the updated data, it calls the generation AI model means again to generate a new product candidate list. This new candidate list is also shaped by the above process and sent to the smart glasses again.

[0547] Step 9:

[0548] The smart glasses will then display the new product candidate list using a visualization tool and present it to the user. The user will then review the products based on the information provided and make a final selection. After making their selection, they will be guided through the purchase process.

[0549] Through these steps, users can efficiently select the best mattress for them. This system provides personalized product recommendations based on the user's emotional state and efficient in-store navigation.

[0550] 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.

[0551] 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.

[0552] 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.

[0553] [Second embodiment]

[0554] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0555] 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.

[0556] 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).

[0557] 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.

[0558] 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.

[0559] 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).

[0560] 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.

[0561] 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.

[0562] 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.

[0563] 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.

[0564] 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.

[0565] 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."

[0566] The present invention provides a system that reduces the time and effort required for a user to select an optimal product, particularly a mattress. Specific embodiments of the system are described below.

[0567] Overall system overview

[0568] In this system, users input criteria data via a chat-style interface, and the server uses a generative AI model to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list using the generative AI model again based on that feedback.

[0569] User Action

[0570] Enter the condition:

[0571] Users access a chat-style input screen using their device and enter their desired conditions, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[0572] Server Processing

[0573] Receiving user data:

[0574] The server receives the condition data sent by the user. The received data has the following format:

[0575] json

[0576] {

[0577] "budget": "50000",

[0578] "hardness": "medium",

[0579] "Size": "Single",

[0580] "Brand": "Not specified"

[0581] }

[0582] Invoke the generative AI model:

[0583] The server passes the received condition data to the generative AI model, which analyzes the big data. The generative AI model retrieves relevant mattress data from the database and generates a list of optimal product candidates that match the conditions.

[0584] Generate candidate list:

[0585] Based on the information obtained from the AI ​​model, the server formats the product candidate list and converts it into a format that can be sent back to the user. For example, the following candidate list is generated:

[0586] json

[0587] [

[0588] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0589] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0590] ]

[0591] Sending results to the user:

[0592] The server returns the formatted candidate list to the user's chat screen.

[0593] Feedback Processing

[0594] User feedback:

[0595] The user can review the provided list of candidates and provide feedback on what is closest to their desired fit or on any additional requirements. For example, they could enter, "Mattress A seems good, but let me know if there is anything cheaper."

[0596] Re-invoke the generated AI model:

[0597] The server receives feedback from the user, reflects the revised conditions, and invokes the generative AI model again. This process generates the optimal candidate list based on the new conditions from the database, just as it did last time.

[0598] Generate a new candidate list:

[0599] Based on the revised conditions, the server again reformats the information obtained from the generative AI model and returns a new product candidate list to the user. For example, the following additional candidates are generated:

[0600] json

[0601] [

[0602] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0603] ]

[0604] Final selection and purchase

[0605] Final decision:

[0606] The user finally selects the best product from the presented options and makes a decision, such as "I will purchase Mattress C."

[0607] Purchase procedure:

[0608] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[0609] Specific examples

[0610] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0611] 2. The server receives the condition data, and the generative AI model uses big data to generate the optimal product candidate list.

[0612] 3. The server returns the generated candidate list to the user, presenting "Mattress A" and "Mattress B."

[0613] 4. The user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[0614] 5. The server invokes the generative AI model again based on the feedback and generates a new candidate list.

[0615] 6. The server returns a new candidate, "Mattress C," to the user.

[0616] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[0617] In this way, the system allows users to efficiently find the best mattress.

[0618] The processing flow will be explained below.

[0619] Step 1:

[0620] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[0621] Step 2:

[0622] Server: The server receives the condition data sent by the user. The received data is in the following format:

[0623] json

[0624] {

[0625] "budget": "50000",

[0626] "hardness": "medium",

[0627] "Size": "Single",

[0628] "Brand": "Not specified"

[0629] }

[0630] Step 3:

[0631] Server: The server passes the received condition data to the generative AI model and invokes it. The generative AI model analyzes the condition data using big data and generates an optimal product candidate list.

[0632] Step 4:

[0633] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[0634] Step 5:

[0635] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[0636] json

[0637] [

[0638] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0639] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0640] ]

[0641] Step 6:

[0642] Server: The server returns the formatted product candidate list to the user's chat screen.

[0643] Step 7:

[0644] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[0645] Step 8:

[0646] Server: The server receives feedback from the user, reflects the revised conditions, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[0647] Step 9:

[0648] Generative AI model: The generative AI model analyzes the condition data again and generates a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[0649] Step 10:

[0650] Server: The server formats the new product list and sends it back to the user, for example:

[0651] json

[0652] [

[0653] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0654] ]

[0655] Step 11:

[0656] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[0657] Step 12:

[0658] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[0659] Example 1

[0660] 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."

[0661] In the past, when users selected products such as mattresses, it took a lot of time and effort to find the best product from the vast number of options available. Furthermore, manually searching and comparing product information was cumbersome, making it difficult for users to find the product that best met their requirements.

[0662] 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.

[0663] In this invention, the server includes means for inputting user condition data, data processing means for receiving the user condition data, machine learning model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, data processing means for processing the generated product candidate list and returning it to the user, and data processing means for receiving feedback from the user, reflecting condition modifications, and again calling the machine learning model means, thereby enabling users to efficiently find products that best fit their conditions.

[0664] "User condition data" refers to the desired conditions that a user inputs to select a specific product, and includes data such as budget, hardness, size, brand, etc.

[0665] "Means for inputting user condition data" refers to devices or software that provide an interface for users to input their desired conditions. This includes smartphones, tablets, PC applications, and web browsers.

[0666] "Data processing means for receiving user condition data" refers to a server or network device for receiving and processing the condition data sent by the user.

[0667] "Big data" refers to a huge amount of data, a dataset that can be analyzed to extract valuable information, such as product information, reviews, and pricing data.

[0668] A "machine learning model" is an algorithm or program used to analyze data and generate an optimal list of product candidates based on the results. A specific example is a generative AI model.

[0669] The "generative machine learning model means" refers to software and hardware for receiving user condition data and large-scale data as input and executing a machine learning model to generate an optimal product candidate list.

[0670] "Data processing means" refers to the server and related software for receiving, processing, transmitting and feedback processing data.

[0671] "Product candidate list" refers to a list of product candidates generated based on the user's criteria and presented to the user for consideration.

[0672] "Feedback" refers to the user inputting their opinions and further requirements regarding the provided product candidate list.

[0673] "Condition modification" refers to changing or adding to the original condition data based on user feedback.

[0674] "Invoke" refers to launching a specific function or process (in this case, a machine learning model) to execute it.

[0675] The present invention provides a system for efficiently selecting a desired product, particularly a mattress, by a user. Specific embodiments will be described below.

[0676] Hardware and Software Use

[0677] User condition data input

[0678] The user accesses a chat-style input screen using a web browser or dedicated application on a device (smartphone, tablet, PC, etc.) and enters desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0679] Receiving and processing condition data

[0680] The server receives the condition data sent by the user as an HTTP POST request and parses it in JSON format. The data can be processed using a web server such as NGINX or Apache, or a programming language such as Python or Node.js. The received data is in the following format:

[0681] json

[0682] {

[0683] "budget": "50000",

[0684] "hardness": "medium",

[0685] "Size": "Single",

[0686] "Brand": "Not specified"

[0687] }

[0688] Invoking a generative AI model

[0689] The server invokes a generative AI model (e.g., OpenAI's GPT-3 or Google's BERT) based on the received condition data. The prompt is created as follows:

[0690] "Choose the best mattress for you, with a budget of under ¥50,000, medium firmness, single size, and no particular brand preference."

[0691] Generate product candidate list

[0692] The generative AI model analyzes large-scale data and generates a list of product candidates that match the criteria. The generated list of product candidates is received by the server and formatted to be sent back to the user. Examples of formatting are as follows:

[0693] "Mattress A: Price 48,000 yen, medium firmness, size single, brand X"

[0694] "Mattress B: Price 45,000 yen, medium firmness, size single, brand Y"

[0695] Sending results to the user

[0696] The server returns the formatted product candidate list to the user's chat screen, and the user confirms the provided list.

[0697] Feedback Processing

[0698] The user reviews the provided candidate list and provides feedback on what is closest to their desired outcome or on additional requirements. For example, they might input, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback, reflects the revised requirements, and invokes the generative AI model again. The new prompt text becomes:

[0699] "Choose the best mattress that is medium firm, single size, and doesn't specify the brand, with a budget of 45,000 yen or less, if possible."

[0700] The newly generated product candidate list is formatted in the same way and sent back to the user.

[0701] Specific examples

[0702] When a user inputs "budget under ¥50,000, medium firmness, single size, no particular brand," the server receives the condition data and the generative AI model uses big data to generate an optimal product candidate list. On the other hand, if the user inputs feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," the server readjusts the conditions and sends a new prompt to the generative AI model, generating a new product candidate list.

[0703] In this way, the present invention is a system that efficiently generates optimal product candidates based on condition data, significantly reducing the time and effort required by the user.

[0704] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0705] Step 1:

[0706] The user accesses a chat-style input screen using a device (smartphone, tablet, PC, etc.) and enters the desired conditions. For example, they might enter "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0707] Input: User's desired conditions (budget, hardness, size, brand)

[0708] Output: User's desired conditions data

[0709] Step 2:

[0710] The device sends the user's criteria data to the server, which receives it as an HTTP POST request and parses it into JSON format.

[0711] Input: User's desired conditions data

[0712] Output: Condition data in JSON format

[0713] Step 3:

[0714] The server uses the received condition data to create a prompt for the generative AI model, such as "Please choose the best mattress for me, within a budget of ¥50,000, with medium firmness, single size, and no particular brand preference."

[0715] Input: Condition data in JSON format

[0716] Output: prompt statement

[0717] Step 4:

[0718] The server sends the prompt to a generative AI model (such as OpenAI's GPT-3 or BERT), which analyzes the prompt and generates an optimal product candidate list from large-scale data.

[0719] Input: prompt statement

[0720] Output: Raw product candidate list data

[0721] Step 5:

[0722] The server receives the raw data of the product candidate list returned by the generative AI model and formats it into a format that is easy for users to understand, such as "Mattress A: Price 48,000 yen, Medium firmness, Size single, Brand X."

[0723] Input: Raw data of product candidate list

[0724] Output: A formatted list of product candidates

[0725] Step 6:

[0726] The server returns the formatted product candidate list to the user's chat screen, which the user can confirm.

[0727] Input: A formatted list of product candidates

[0728] Output: A list of product candidates presented to the user

[0729] Step 7:

[0730] The user enters feedback on the presented product candidate list. For example, "Mattress A seems good, but please let me know if there is anything cheaper."

[0731] Input: User feedback

[0732] Output: Feedback data

[0733] Step 8:

[0734] The device sends the user's feedback data to the server. The server receives this data and modifies the condition data. It then creates a new prompt and sends it back to the generative AI model. For example, it generates a prompt such as, "Please choose the best mattress for me, preferably within 45,000 yen, with medium firmness, single size, and no particular brand preference."

[0735] Input: Feedback data

[0736] Output: Modified prompt statement

[0737] Step 9:

[0738] The generative AI model generates a new product candidate list based on the new prompt. The server receives this and formats it again. The formatted list is then sent to the user again.

[0739] Input: modified prompt text

[0740] Output: Regenerated product candidate list

[0741] Step 10:

[0742] Finally, the user selects the most suitable product from the presented product candidates and notifies the system of their intention to purchase the selected product. The server provides the user with information regarding the purchase procedure.

[0743] Input: Final selected product data

[0744] Output: Purchase procedure guide information

[0745] (Application example 1)

[0746] 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."

[0747] Current online shopping sites have the problem that it is difficult for users to efficiently find the products they want. In particular, because they do not provide prompt responses to input of conditions or feedback, users must spend a great deal of time and effort to find the optimal product. The purpose of this invention is to solve these problems and provide a system that allows users to efficiently select the optimal product.

[0748] 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.

[0749] In this invention, the server includes means for inputting user condition data, means for receiving the user condition data, a generation AI model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, means for formatting the generated product candidate list and returning it to the user, means for receiving feedback from the user and reflecting the condition modifications to call the generation AI model means again, a user interface means using a smartphone, and means for presenting the generated product candidate list to the user and providing feedback on additional conditions, thereby enabling the user to quickly and efficiently select optimal products.

[0750] "User condition data" refers to the desired conditions entered by the user when selecting a product, such as budget, size, hardness, brand, and other specific requirements.

[0751] "Input means" refers to the interface or device through which the user inputs condition data, such as a smartphone, PC, or tablet.

[0752] "Server means" refers to a computer system for receiving, storing, and analyzing data sent by users, including databases and cloud servers.

[0753] "Generative AI model means" refers to an artificial intelligence model used to analyze received data and generate optimal product candidates, including, for example, a machine learning model or a natural language processing model.

[0754] "Product candidate list" refers to data that lists multiple candidate products generated by a generative AI model based on user condition data.

[0755] "Formatting" refers to the process of converting the generated product candidate list into a format that is easy for users to view and understand.

[0756] "Receiving feedback" refers to the process of receiving additional conditions, opinions, etc. from users via input devices.

[0757] "Condition Modification" refers to the process of changing or modifying the original condition data based on feedback received from users.

[0758] "Invoking the generative AI model means again" refers to the process of generating a new product candidate list using the generative AI model again based on the modified condition data.

[0759] "Smartphone-based user interface means" refers to an interface that allows a user to interact with the system via a smartphone, input conditions, and send feedback.

[0760] "Means for providing feedback on additional conditions" refers to a process or interface that allows users to input additional requirements or requested changes and send them to the server.

[0761] System Overview

[0762] The present invention provides a system that allows users to select optimal products, particularly through a smartphone-based user interface, enabling them to find products efficiently and quickly. This system inputs user criteria data and generates an optimal product candidate list using a generative AI model based on that data. The system then presents the generated product candidate list to the user, receives feedback, and uses the generative AI model again to provide the optimal products that meet the user's needs.

[0763] Hardware and Software

[0764] The user's smartphone acts as the input interface, and the server receives the data, runs the generative AI model, generates a list of product candidates, and processes the feedback. For generative AI models, OpenAI GPT-4 and Google BERT are used as examples.

[0765] Program processing

[0766] 1. Enter your user data:

[0767] Users use their smartphones to access the chatbot screen on a dedicated application.

[0768] Users can input the necessary criteria for product selection via text or voice, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[0769] 2. Sending and receiving condition data:

[0770] The smartphone application converts the condition data entered by the user into JSON format and sends it to the server.

[0771] The server analyzes the received condition data and inputs it into the generative AI model.

[0772] 3. Run the generative AI model:

[0773] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[0774] The generative AI model searches and analyzes relevant product information from an internal database to extract the most suitable candidates.

[0775] 4. Generate and return product candidate list:

[0776] The server formats the generated product candidate list and converts it into a format that is easy for the user to understand.

[0777] The formatted product candidate list is sent back to the smartphone application and displayed on the user's screen.

[0778] Specific examples

[0779] If a user types "I'm looking for a medium-firm mattress under 50,000 yen" into a smartphone app, the system will operate as follows:

[0780] 1. A user types into a smartphone app, "I'm looking for a medium-firm mattress under 50,000 yen."

[0781] 2. This text data is converted to JSON format and sent to the server.

[0782] 3. The server inputs the data into a generative AI model (e.g., OpenAI GPT-4 or Google BERT) to generate a list of product candidates that match the criteria.

[0783] 4. Format the generated product candidate list and return it to the user in the following format:

[0784] "Mattress A, Price: 48,000 yen, Firmness: Medium, Size: Single, Brand: Brand X"

[0785] "Mattress B, Price: 45,000 yen, Firmness: Medium, Size: Single, Brand: Brand Y"

[0786] Prompt Sentence Examples

[0787] An example of a prompt sentence to input to a generative AI model is as follows.

[0788] "I'm looking for a mattress. My budget is under ¥50,000, I want a medium-firm mattress, and I don't have a specific brand in mind. Can you recommend the best product for me?"

[0789] In this way, it is possible to realize a system that allows users to quickly and efficiently select the most suitable product.

[0790] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0791] Step 1:

[0792] The user accesses the chatbot screen of the dedicated application using a smartphone. The user inputs the necessary product selection criteria by text or voice. For example, "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0793] Input: The condition data entered by the user.

[0794] Output: Condition data collected on the user's device.

[0795] Step 2:

[0796] The terminal converts the condition data entered by the user into JSON format and sends it to the server.

[0797] Input: Condition data.

[0798] Data processing: Conversion to JSON format.

[0799] Output: Criteria data converted to JSON format.

[0800] Step 3:

[0801] The server analyzes the received JSON formatted condition data and inputs it into the generative AI model.

[0802] Input: Criteria data in JSON format.

[0803] Data analysis: Parsing JSON data and applying it to generative AI models.

[0804] Output: Analysis results and inputs to generative AI models.

[0805] Step 4:

[0806] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[0807] Input: Parsed condition data.

[0808] Data computation: Generative AI models generate product candidate lists.

[0809] Output: Product candidate list.

[0810] Step 5:

[0811] The generated product candidate list is formatted on the server and converted into a format that is easy for users to understand.

[0812] Input: Product picklist.

[0813] Data processing: Formatting the product candidate list.

[0814] Output: A formatted list of product suggestions.

[0815] Step 6:

[0816] The server returns the formatted product candidate list to the user's smartphone application, which displays it on the user's screen.

[0817] Input: A formatted list of product suggestions.

[0818] Output: Product candidate list displayed on user's device.

[0819] Step 7:

[0820] The user checks the product candidate list, sets additional requirements and feedback, and re-enters the list. An example of feedback is, "Mattress A is good, but I want something a little cheaper."

[0821] Input: User feedback.

[0822] Output: Feedback data with added and modified conditions.

[0823] Step 8:

[0824] The server again receives feedback from the user and generates new JSON data that reflects the modified conditions.

[0825] Input: Feedback data.

[0826] Data processing: Generate JSON data that reflects the condition modifications.

[0827] Output: The modified JSON data.

[0828] Step 9:

[0829] Based on the revised condition data, the server calls the generative AI model again and generates a new list of product candidates.

[0830] Input: The modified condition data.

[0831] Data computation: Generative AI models generate new product candidate lists.

[0832] Output: A new list of product candidates.

[0833] Step 10:

[0834] The server then formats the regenerated product candidate list, converts it into a format that is easy for the user to understand, and returns it to the user's smartphone application.

[0835] Input: A new product shortlist.

[0836] Data processing: Formatting a new product candidate list.

[0837] Output: The formatted new product candidate list displayed on the user's device.

[0838] This process creates a system that allows users to efficiently select the most suitable product.

[0839] 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.

[0840] The present invention provides a system that reduces the time and effort required for users to select optimal products, especially mattresses. A key feature of the system is the incorporation of an emotion engine that recognizes the user's emotions and selects products that reflect their emotional state, enabling more personalized recommendations. Specific embodiments of the system are described below.

[0841] Overall system overview

[0842] In this system, users input criteria data via a chat-style interface, and the server then uses a generative AI model and an emotion engine to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[0843] User Action

[0844] Enter the condition:

[0845] Users access a chat-style input screen using their device and input their desired conditions (budget, firmness, size, brand, etc.) For example, they could input "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0846] Server Processing

[0847] Receiving user data:

[0848] The server receives the condition data sent by the user. The received data has the following format:

[0849] json

[0850] {

[0851] "budget": "50000",

[0852] "hardness": "medium",

[0853] "Size": "Single",

[0854] "Brand": "Not specified"

[0855] }

[0856] Calling the Emotion Engine:

[0857] Based on the received condition data, the server calls the emotion engine and analyzes the user's emotional state. The emotion engine recognizes emotions through chat context, user input, facial recognition, and voice analysis.

[0858] Processing generative AI models

[0859] Invoke the generative AI model:

[0860] The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[0861] Generate product picklist:

[0862] The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria and emotional state. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[0863] Server Processing

[0864] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[0865] json

[0866] [

[0867] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0868] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0869] ]

[0870] Sending results to the user:

[0871] The server returns the formatted product candidate list to the user's chat screen.

[0872] Feedback Processing

[0873] User feedback:

[0874] The user can review the provided list of product candidates and provide feedback on what is closest to their desired product or on any additional requirements. For example, they can enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[0875] Re-invoke the generated AI model:

[0876] The server receives feedback from the user, reflects the modified conditions and emotional state, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[0877] Regenerate product pick list

[0878] Generative AI models:

[0879] The generative AI model then analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[0880] server:

[0881] The server then formats the new product list and sends it back to the user, for example:

[0882] json

[0883] [

[0884] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0885] ]

[0886] Final selection and purchase

[0887] Final decision:

[0888] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[0889] Purchase procedure:

[0890] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details of the process.

[0891] Specific examples

[0892] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0893] 2. The server receives the condition data, and the emotion engine recognizes the user's emotional state. At this time, it determines that the user is feeling stressed.

[0894] 3. The generative AI model uses big data to analyze the conditions and emotional state and suggests mattresses with relaxing effects.

[0895] 4. The server returns the generated candidate lists "Mattress A" and "Mattress B" to the user.

[0896] 5. A user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[0897] 6. The server passes the feedback and emotional state back to the generative AI model and suggests a new candidate, "Mattress C."

[0898] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[0899] The system allows users to find the perfect mattress in an efficient and emotionally sensitive manner.

[0900] The processing flow will be explained below.

[0901] Step 1:

[0902] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[0903] Step 2:

[0904] Server: The server receives the condition data sent by the user. The received data is in the following format:

[0905] json

[0906] {

[0907] "budget": "50000",

[0908] "hardness": "medium",

[0909] "Size": "Single",

[0910] "Brand": "Not specified"

[0911] }

[0912] Step 3:

[0913] Server: Based on the received condition data, the server calls the emotion engine to analyze the user's emotional state. The emotion engine recognizes the user's emotions using the user's input and other emotional indicators (e.g., context, facial recognition, voice analysis, etc.).

[0914] Step 4:

[0915] Emotion engine: The emotion engine recognizes the user's emotional state, for example, determining that the user is feeling stressed. The recognized emotional state is returned to the server.

[0916] Step 5:

[0917] Server: The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[0918] Step 6:

[0919] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the user's criteria and emotional state. This process involves running database queries, filtering, and finally ranking the products by scoring.

[0920] Step 7:

[0921] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be returned to the user. For example, the following candidate list is generated:

[0922] json

[0923] [

[0924] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[0925] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[0926] ]

[0927] Step 8:

[0928] Server: The server returns the formatted product candidate list to the user's chat screen.

[0929] Step 9:

[0930] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[0931] Step 10:

[0932] Server: The server receives feedback from the user and passes it back to the generative AI model along with condition modifications and emotional states.

[0933] Step 11:

[0934] Generative AI model: The generative AI model analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[0935] Step 12:

[0936] Server: The server formats the new product list and sends it back to the user, for example:

[0937] json

[0938] [

[0939] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[0940] ]

[0941] Step 13:

[0942] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[0943] Step 14:

[0944] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[0945] Example 2

[0946] 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."

[0947] Conventional product recommendation systems only suggest products based on the user's criteria, without taking into account the user's emotional state. This creates a problem in that they provide products without understanding the user's feelings when searching for a product. As a result, the suggested products often do not match the user's actual needs, making it difficult to increase satisfaction.

[0948] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an input device means for inputting user condition data, a server device means for receiving the user condition data, a generation AI model means for generating an optimal product candidate list based on the received condition data and the user's emotional state recognized by the emotion analysis means, a server device means for shaping the generated product candidate list and returning it to the user, and a server device means for receiving feedback from the user, reflecting the condition modifications and emotional state, and again calling the generation AI model means. This enables personalized product recommendations that take the user's emotional state into consideration.

[0949] "User" means a person who uses the System to select and purchase a Product.

[0950] "Condition data" is information about desired conditions (budget, hardness, size, brand, etc.) that a user enters into the system.

[0951] The "input device means" is a device (such as a personal computer or smartphone) that the user uses to input condition data.

[0952] "Server device means" refers to a server device for receiving and processing data from a user.

[0953] "Emotion analysis means" refers to technology that recognizes the user's emotional state through user input, facial recognition, voice analysis, etc.

[0954] The "generative AI model means" is an artificial intelligence model that analyzes big data based on the user's condition data and emotional state, and generates an optimal product candidate list.

[0955] A "product candidate list" is a plurality of candidate products selected by a generative AI model means and suggested to a user.

[0956] "Feedback" refers to the act of a user providing more desirable or additional conditions for a product candidate list.

[0957] "Condition modification" is the process of changing and updating the initial conditions based on user feedback.

[0958] "Invoking the generating AI model means again" is a procedure for causing the generating AI model means to generate a product candidate list again based on the condition modification and emotional state.

[0959] This invention relates to a system that reduces the time and effort required for users to select the optimal product, especially a mattress. This system is capable of providing more personalized recommendations by incorporating an emotion engine that recognizes the user's emotions and selects products that reflect the user's emotional state.

[0960] Overall system overview

[0961] In this system, users input criteria data through a chat-style interface, and the server generates and provides an optimal product candidate list using a generative AI model and emotion engine based on that data. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[0962] Specific Embodiments

[0963] User input operations

[0964] The user accesses the chat screen of the system using a terminal and inputs the desired data. For example, the user can input specific desired data such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference." This input information is input in text format as follows:

[0965] Budget: 50,000 yen or less, medium hardness, single size, no particular brand preference.

[0966] Server reception and analysis

[0967] The server receives the condition data sent by the user and analyzes it. The data is sent to the server in JSON format, and then the emotion engine is called to analyze the user's emotional state. The emotion engine uses the device's camera and microphone to perform facial recognition and voice analysis, and obtains an analysis result such as "the user is feeling stressed."

[0968] Invoking a generative AI model

[0969] The server passes the analysis results and condition data to the AI ​​model to generate the optimal product candidate list. The input prompt is set as follows:

[0970] Budget: ¥50,000 or less, medium hardness, single size, no particular brand. Users are feeling stressed.

[0971] The generative AI model analyzes large amounts of data and selects the best product candidates based on criteria and emotional state.

[0972] Generate and provide product candidate lists

[0973] The product candidate list generated by the generative AI model is formatted on the server and provided to the user. Specifically, the server replies to the user's chat screen with a message like the following:

[0974] The recommended mattresses are as follows:

[0975] 1. Mattress A - Price: 48,000 yen - Firmness: Medium - Size: Single - Brand X

[0976] 2. Mattress B - Price: 45,000 yen - Firmness: Medium - Size: Single - Brand Y

[0977] Receiving feedback and reanalyzing

[0978] The user reviews the provided product candidate list and provides feedback with more detailed requirements. For example, they may send feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback and calls the generative AI model again to begin the process of generating a new product candidate list.

[0979] Providing a regenerated candidate list

[0980] The server again formats the new product candidate list obtained from the generative AI model and provides it to the user in the following format:

[0981] This is a new proposal.

[0982] 1. Mattress C - Price: 42,000 yen - Firmness: Medium - Size: Single - Brand: Z

[0983] Final selection and checkout

[0984] The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C." The server receives the selection and guides the user to detailed information about the purchase procedure and a link to the purchase page.

[0985] This system allows users to efficiently select the most suitable product while taking into account their own emotional state.

[0986] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0987] Step 1: User Input

[0988] Users access the chat screen of the system using a terminal and input their desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[0989] Input: User's desired conditions (budget, hardness, size, brand, etc.)

[0990] Output: The user's desired conditions are sent from the terminal in text format.

[0991] Step 2: Server Receive

[0992] The server receives the condition data sent by the user and prepares to analyze the received data.

[0993] Input: The user's desired conditions are sent to the server in text format.

[0994] Output: The criteria data is converted to JSON format and passed to the next processing step.

[0995] Step 3: Sentiment Analysis

[0996] The server calls the emotion engine based on the condition data and analyzes the user's emotional state. The emotion engine uses the device's camera and microphone to recognize the user's face and analyze their voice.

[0997] Input: Condition data (JSON format) and user input

[0998] Output: Sentiment analysis result (e.g., user is stressed)

[0999] Step 4: Call the generative AI model

[1000] The server passes the analyzed emotional state and condition data to a generative AI model, which then generates an optimal product candidate list. The generative AI model performs analysis using big data.

[1001] Input: Condition data and emotion analysis results

[1002] Output: Product candidate list generated by the generative AI model

[1003] Step 5: Product candidate list formatting

[1004] The server formats the product candidate list obtained from the generative AI model and prepares it for return to the user.

[1005] Input: Product candidate list generated by the generative AI model

[1006] Output: A formatted list of product candidates

[1007] Step 6: Reply to the user

[1008] The server sends the formatted product candidate list to the user's chat screen.

[1009] Input: A formatted list of product candidates

[1010] Output: Product candidate list displayed on the user's chat screen

[1011] Step 7: User feedback

[1012] Users can review the provided product shortlist and enter additional preferences and feedback, such as "Mattress A looks good, but let me know if there's anything cheaper."

[1013] Input: User additional requirements and feedback

[1014] Output: User feedback sent from the terminal in text format

[1015] Step 8: Server reception and reparsing

[1016] The server receives feedback sent by the user and invokes a regenerative AI model that reflects condition modifications and emotional states.

[1017] Input: User feedback (text format)

[1018] Output: Reanalyzed condition data and sentiment analysis results

[1019] Step 9: Regenerate AI model call

[1020] The server then calls the generative AI model again based on the reanalyzed condition data and emotional state, and generates a new list of product candidates.

[1021] Input: Reanalyzed condition data and emotional state

[1022] Output: A new product candidate list generated by the generative AI model

[1023] Step 10: Provide a regenerated candidate list

[1024] The server formats the regenerated product candidate list and provides it to the user's chat screen again.

[1025] Input: A new product candidate list generated by the regenerative AI model

[1026] Output: New, re-formatted product candidate list

[1027] Step 11: Final user selection

[1028] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[1029] Input: User's final selection

[1030] Output: The final selected product information is sent to the server.

[1031] Step 12: Purchase procedure guide

[1032] The server receives the final selected product information and provides the user with detailed purchase procedure information and a link to the purchase page.

[1033] Input: User's last selected product information

[1034] Output: A purchase procedure message will be displayed on the user's chat screen.

[1035] By combining the above steps, users can efficiently select the most suitable mattress and proceed with the purchase process while taking into account their own emotional state.

[1036] (Application example 2)

[1037] 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."

[1038] Conventional product selection systems suggest products based on user condition data, but because they do not take the user's emotional state into account, they have the problem of not being able to make suggestions that are optimal for the user's psychological state. In addition, there is no way to display product location information within the store, making it difficult for users to search for products efficiently. This reduces users' motivation to purchase, and there are issues such as the time and effort required to select the most suitable product.

[1039] 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: an input means for inputting user condition data; a communication means for receiving the user condition data; a generation AI model means for analyzing the received condition data and emotion analysis data to generate an optimal product candidate list; a communication means for formatting the generated product candidate list and returning it to the user; a communication means for reflecting condition modifications based on user feedback and emotion analysis data and again calling the generation AI model means; an emotion recognition means for recognizing the user's emotional state in real time; and a visualization means for displaying product locations within a physical store for the user. This enables personalized product suggestions based on the user's emotional state, allowing the user to efficiently search for products within the store.

[1040] "User condition data" refers to the desired conditions (budget, hardness, size, brand, etc.) entered by the user.

[1041] "Input means" refers to a device or interface that allows a user to input condition data.

[1042] "Communication means" refers to a network interface for sending and receiving data.

[1043] "Emotion analysis data" is data that represents the user's emotional state after analyzing it.

[1044] "Generative AI model means" refers to an artificial intelligence model for generating an optimal product candidate list using received data.

[1045] "Emotion recognition means" refers to technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state in real time.

[1046] "Visualization means" refers to displays and AR technologies that allow users to visually view information.

[1047] "Product Candidate List" means a list of optimal products generated by the generative AI model means and based on the user's criteria and emotional state.

[1048] "Feedback" refers to any additional preferences or comments provided by the User.

[1049] "Physical store" refers to an actual store space.

[1050] To implement this invention, a system including the following components is required. This system is designed to suggest optimal products based on a user's condition data and emotion data. Specifically, it uses a user input means, a communication means, an emotion recognition means, a generation AI model means, and a visualization means.

[1051] First, the user puts on the smart glasses and inputs their requirements by voice or touch. This includes information such as budget, hardness, size, and brand. For example, "budget is under ¥50,000, hardness is medium, size is single, and brand is not specified."

[1052] The input condition data is sent to the server via a communication means. The server receives the data and analyzes the user's emotional state via an emotion recognition means. Emotion recognition uses the smart glasses' camera and microphone to analyze the user's facial expressions and tone of voice. For example, it can determine whether the user is feeling stressed.

[1053] The server then passes the received condition data and emotion data to the generative AI model means, which analyzes this information using big data. The generative AI model generates an optimal product candidate list based on the condition and emotion state. The generated list is sent back to the smart glasses via the communication means.

[1054] The smart glasses use a visualization tool to present the generated product candidate list to the user, along with the product location information, allowing the user to efficiently search for products in the store.

[1055] For example, if "Mattress A" is included in the generated candidate list, its location information will be displayed on the display. The user can walk around the store while checking this information and check the products directly. If the user gives feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," this feedback will also be sent to the server via the communication means.

[1056] The server then calls the generative AI model again based on the received feedback and emotional data, updating the product candidate list. In this way, the user's feedback and emotional state are reflected in real time, allowing for smooth selection of the most suitable product.

[1057] As a concrete example, the following prompt sentence is input to the generative AI model:

[1058] "List products (mattresses) that match the following criteria: Conditions: { "Budget": "50000", "Firmness": "Medium", "Size": "Single", "Brand": "Not Specified"} User emotional state: "Feeling stressed""

[1059] The system allows users to find the perfect mattress in an efficient and emotionally sensitive way.

[1060] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1061] Step 1:

[1062] The user puts on the smart glasses and inputs the desired conditions by voice or touch operation. The input conditions include budget, hardness, size, brand, etc. Data such as "budget within 50,000 yen, hardness medium, size single, no particular brand" is input. The input data is temporarily stored in the smart glasses' internal memory.

[1063] Step 2:

[1064] The user's input data is sent to the server via a communication method, either via the internet or the store's Wi-Fi. The server processes the received condition data as data for analysis. At this time, the received data is converted to JSON format and saved in the server's data storage.

[1065] Step 3:

[1066] The server then calls the emotion recognition unit based on the received data. It uses the smart glasses' camera and microphone to collect the user's facial expressions and tone of voice, analyzes them, and recognizes their emotional state. The analyzed emotional state is then classified into categories such as "feeling stressed" and output as data.

[1067] Step 4:

[1068] The server passes the condition data and the emotion data obtained by emotion recognition to the generative AI model means. The generative AI model receives these data as input and performs analysis using big data. The analysis process includes running database queries, filtering, and scoring. The generative AI model outputs a list of optimal product candidates, which is sent back to the server in JSON format.

[1069] Step 5:

[1070] The server then formats the received product candidate list and converts it into a data format for sending back to the user. This conversion process includes formatting data such as product name, price, and location information. The formatted data is then sent back to the smart glasses via a communication means.

[1071] Step 6:

[1072] The smart glasses display the received product candidate list to the user using a visualization means. The smart glasses display shows the names, prices, and location information of the candidate products in the store. The user confirms the displayed information and moves around the store to actually check the products.

[1073] Step 7:

[1074] After checking the products, the user can provide feedback by voice input or touch operation on the smart glasses. For example, they can input feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." This feedback is sent to the server via a communication means.

[1075] Step 8:

[1076] The server receives the feedback and updates the condition data and emotion data. Based on the updated data, it calls the generation AI model means again to generate a new product candidate list. This new candidate list is also shaped by the above process and sent to the smart glasses again.

[1077] Step 9:

[1078] The smart glasses will then display the new product candidate list using a visualization tool and present it to the user. The user will then review the products based on the information provided and make a final selection. After making their selection, they will be guided through the purchase process.

[1079] Through these steps, users can efficiently select the best mattress for them. This system provides personalized product recommendations based on the user's emotional state and efficient in-store navigation.

[1080] 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.

[1081] 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.

[1082] 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.

[1083] [Third embodiment]

[1084] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1085] 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.

[1086] 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).

[1087] 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.

[1088] 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.

[1089] 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).

[1090] 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.

[1091] 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.

[1092] 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.

[1093] 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.

[1094] 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.

[1095] 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."

[1096] The present invention provides a system that reduces the time and effort required for a user to select an optimal product, particularly a mattress. Specific embodiments of the system are described below.

[1097] Overall system overview

[1098] In this system, users input criteria data via a chat-style interface, and the server uses a generative AI model to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list using the generative AI model again based on that feedback.

[1099] User Action

[1100] Enter the condition:

[1101] Users access a chat-style input screen using their device and enter their desired conditions, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[1102] Server Processing

[1103] Receiving user data:

[1104] The server receives the condition data sent by the user. The received data has the following format:

[1105] json

[1106] {

[1107] "budget": "50000",

[1108] "hardness": "medium",

[1109] "Size": "Single",

[1110] "Brand": "Not specified"

[1111] }

[1112] Invoke the generative AI model:

[1113] The server passes the received condition data to the generative AI model, which analyzes the big data. The generative AI model retrieves relevant mattress data from the database and generates a list of optimal product candidates that match the conditions.

[1114] Generate candidate list:

[1115] Based on the information obtained from the AI ​​model, the server formats the product candidate list and converts it into a format that can be sent back to the user. For example, the following candidate list is generated:

[1116] json

[1117] [

[1118] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1119] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1120] ]

[1121] Sending results to the user:

[1122] The server returns the formatted candidate list to the user's chat screen.

[1123] Feedback Processing

[1124] User feedback:

[1125] The user can review the provided list of candidates and provide feedback on what is closest to their desired fit or on any additional requirements. For example, they could enter, "Mattress A seems good, but let me know if there is anything cheaper."

[1126] Re-invoke the generated AI model:

[1127] The server receives feedback from the user, reflects the revised conditions, and invokes the generative AI model again. This process generates the optimal candidate list based on the new conditions from the database, just as it did last time.

[1128] Generate a new candidate list:

[1129] Based on the revised conditions, the server again reformats the information obtained from the generative AI model and returns a new product candidate list to the user. For example, the following additional candidates are generated:

[1130] json

[1131] [

[1132] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[1133] ]

[1134] Final selection and purchase

[1135] Final decision:

[1136] The user finally selects the best product from the presented options and makes a decision, such as "I will purchase Mattress C."

[1137] Purchase procedure:

[1138] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[1139] Specific examples

[1140] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1141] 2. The server receives the condition data, and the generative AI model uses big data to generate the optimal product candidate list.

[1142] 3. The server returns the generated candidate list to the user, presenting "Mattress A" and "Mattress B."

[1143] 4. The user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[1144] 5. The server invokes the generative AI model again based on the feedback and generates a new candidate list.

[1145] 6. The server returns a new candidate, "Mattress C," to the user.

[1146] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[1147] In this way, the system allows users to efficiently find the best mattress.

[1148] The processing flow will be explained below.

[1149] Step 1:

[1150] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[1151] Step 2:

[1152] Server: The server receives the condition data sent by the user. The received data is in the following format:

[1153] json

[1154] {

[1155] "budget": "50000",

[1156] "hardness": "medium",

[1157] "Size": "Single",

[1158] "Brand": "Not specified"

[1159] }

[1160] Step 3:

[1161] Server: The server passes the received condition data to the generative AI model and invokes it. The generative AI model analyzes the condition data using big data and generates an optimal product candidate list.

[1162] Step 4:

[1163] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[1164] Step 5:

[1165] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[1166] json

[1167] [

[1168] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1169] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1170] ]

[1171] Step 6:

[1172] Server: The server returns the formatted product candidate list to the user's chat screen.

[1173] Step 7:

[1174] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[1175] Step 8:

[1176] Server: The server receives feedback from the user, reflects the revised conditions, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[1177] Step 9:

[1178] Generative AI model: The generative AI model analyzes the condition data again and generates a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[1179] Step 10:

[1180] Server: The server formats the new product list and sends it back to the user, for example:

[1181] json

[1182] [

[1183] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[1184] ]

[1185] Step 11:

[1186] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[1187] Step 12:

[1188] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[1189] Example 1

[1190] 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."

[1191] In the past, when users selected products such as mattresses, it took a lot of time and effort to find the best product from the vast number of options available. Furthermore, manually searching and comparing product information was cumbersome, making it difficult for users to find the product that best met their requirements.

[1192] 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.

[1193] In this invention, the server includes means for inputting user condition data, data processing means for receiving the user condition data, machine learning model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, data processing means for processing the generated product candidate list and returning it to the user, and data processing means for receiving feedback from the user, reflecting condition modifications, and again calling the machine learning model means, thereby enabling users to efficiently find products that best fit their conditions.

[1194] "User condition data" refers to the desired conditions that a user inputs to select a specific product, and includes data such as budget, hardness, size, brand, etc.

[1195] "Means for inputting user condition data" refers to devices or software that provide an interface for users to input their desired conditions. This includes smartphones, tablets, PC applications, and web browsers.

[1196] "Data processing means for receiving user condition data" refers to a server or network device for receiving and processing the condition data sent by the user.

[1197] "Big data" refers to a huge amount of data, a dataset that can be analyzed to extract valuable information, such as product information, reviews, and pricing data.

[1198] A "machine learning model" is an algorithm or program used to analyze data and generate an optimal list of product candidates based on the results. A specific example is a generative AI model.

[1199] The "generative machine learning model means" refers to software and hardware for receiving user condition data and large-scale data as input and executing a machine learning model to generate an optimal product candidate list.

[1200] "Data processing means" refers to the server and related software for receiving, processing, transmitting and feedback processing data.

[1201] "Product candidate list" refers to a list of product candidates generated based on the user's criteria and presented to the user for consideration.

[1202] "Feedback" refers to the user inputting their opinions and further requirements regarding the provided product candidate list.

[1203] "Condition modification" refers to changing or adding to the original condition data based on user feedback.

[1204] "Invoke" refers to launching a specific function or process (in this case, a machine learning model) to execute it.

[1205] The present invention provides a system for efficiently selecting a desired product, particularly a mattress, by a user. Specific embodiments will be described below.

[1206] Hardware and Software Use

[1207] User condition data input

[1208] The user accesses a chat-style input screen using a web browser or dedicated application on a device (smartphone, tablet, PC, etc.) and enters desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1209] Receiving and processing condition data

[1210] The server receives the condition data sent by the user as an HTTP POST request and parses it in JSON format. The data can be processed using a web server such as NGINX or Apache, or a programming language such as Python or Node.js. The received data is in the following format:

[1211] json

[1212] {

[1213] "budget": "50000",

[1214] "hardness": "medium",

[1215] "Size": "Single",

[1216] "Brand": "Not specified"

[1217] }

[1218] Invoking a generative AI model

[1219] The server invokes a generative AI model (e.g., OpenAI's GPT-3 or Google's BERT) based on the received condition data. The prompt is created as follows:

[1220] "Choose the best mattress for you, with a budget of under ¥50,000, medium firmness, single size, and no particular brand preference."

[1221] Generate product candidate list

[1222] The generative AI model analyzes large-scale data and generates a list of product candidates that match the criteria. The generated list of product candidates is received by the server and formatted to be sent back to the user. Examples of formatting are as follows:

[1223] "Mattress A: Price 48,000 yen, medium firmness, size single, brand X"

[1224] "Mattress B: Price 45,000 yen, medium firmness, size single, brand Y"

[1225] Sending results to the user

[1226] The server returns the formatted product candidate list to the user's chat screen, and the user confirms the provided list.

[1227] Feedback Processing

[1228] The user reviews the provided candidate list and provides feedback on what is closest to their desired outcome or on additional requirements. For example, they might input, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback, reflects the revised requirements, and invokes the generative AI model again. The new prompt text becomes:

[1229] "Choose the best mattress that is medium firm, single size, and doesn't specify the brand, with a budget of 45,000 yen or less, if possible."

[1230] The newly generated product candidate list is formatted in the same way and sent back to the user.

[1231] Specific examples

[1232] When a user inputs "budget under ¥50,000, medium firmness, single size, no particular brand," the server receives the condition data and the generative AI model uses big data to generate an optimal product candidate list. On the other hand, if the user inputs feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," the server readjusts the conditions and sends a new prompt to the generative AI model, generating a new product candidate list.

[1233] In this way, the present invention is a system that efficiently generates optimal product candidates based on condition data, significantly reducing the time and effort required by the user.

[1234] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1235] Step 1:

[1236] The user accesses a chat-style input screen using a device (smartphone, tablet, PC, etc.) and enters the desired conditions. For example, they might enter "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1237] Input: User's desired conditions (budget, hardness, size, brand)

[1238] Output: User's desired conditions data

[1239] Step 2:

[1240] The device sends the user's criteria data to the server, which receives it as an HTTP POST request and parses it into JSON format.

[1241] Input: User's desired conditions data

[1242] Output: Condition data in JSON format

[1243] Step 3:

[1244] The server uses the received condition data to create a prompt for the generative AI model, such as "Please choose the best mattress for me, within a budget of ¥50,000, with medium firmness, single size, and no particular brand preference."

[1245] Input: Condition data in JSON format

[1246] Output: prompt statement

[1247] Step 4:

[1248] The server sends the prompt to a generative AI model (such as OpenAI's GPT-3 or BERT), which analyzes the prompt and generates an optimal product candidate list from large-scale data.

[1249] Input: prompt statement

[1250] Output: Raw product candidate list data

[1251] Step 5:

[1252] The server receives the raw data of the product candidate list returned by the generative AI model and formats it into a format that is easy for users to understand, such as "Mattress A: Price 48,000 yen, Medium firmness, Size single, Brand X."

[1253] Input: Raw data of product candidate list

[1254] Output: A formatted list of product candidates

[1255] Step 6:

[1256] The server returns the formatted product candidate list to the user's chat screen, which the user can confirm.

[1257] Input: A formatted list of product candidates

[1258] Output: A list of product candidates presented to the user

[1259] Step 7:

[1260] The user enters feedback on the presented product candidate list. For example, "Mattress A seems good, but please let me know if there is anything cheaper."

[1261] Input: User feedback

[1262] Output: Feedback data

[1263] Step 8:

[1264] The device sends the user's feedback data to the server. The server receives this data and modifies the condition data. It then creates a new prompt and sends it back to the generative AI model. For example, it generates a prompt such as, "Please choose the best mattress for me, preferably within 45,000 yen, with medium firmness, single size, and no particular brand preference."

[1265] Input: Feedback data

[1266] Output: Modified prompt statement

[1267] Step 9:

[1268] The generative AI model generates a new product candidate list based on the new prompt. The server receives this and formats it again. The formatted list is then sent to the user again.

[1269] Input: modified prompt text

[1270] Output: Regenerated product candidate list

[1271] Step 10:

[1272] Finally, the user selects the most suitable product from the presented product candidates and notifies the system of their intention to purchase the selected product. The server provides the user with information regarding the purchase procedure.

[1273] Input: Final selected product data

[1274] Output: Purchase procedure guide information

[1275] (Application example 1)

[1276] 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."

[1277] Current online shopping sites have the problem that it is difficult for users to efficiently find the products they want. In particular, because they do not provide prompt responses to input of conditions or feedback, users must spend a great deal of time and effort to find the optimal product. The purpose of this invention is to solve these problems and provide a system that allows users to efficiently select the optimal product.

[1278] 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.

[1279] In this invention, the server includes means for inputting user condition data, means for receiving the user condition data, a generation AI model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, means for formatting the generated product candidate list and returning it to the user, means for receiving feedback from the user and reflecting the condition modifications to call the generation AI model means again, a user interface means using a smartphone, and means for presenting the generated product candidate list to the user and providing feedback on additional conditions, thereby enabling the user to quickly and efficiently select optimal products.

[1280] "User condition data" refers to the desired conditions entered by the user when selecting a product, such as budget, size, hardness, brand, and other specific requirements.

[1281] "Input means" refers to the interface or device through which the user inputs condition data, such as a smartphone, PC, or tablet.

[1282] "Server means" refers to a computer system for receiving, storing, and analyzing data sent by users, including databases and cloud servers.

[1283] "Generative AI model means" refers to an artificial intelligence model used to analyze received data and generate optimal product candidates, including, for example, a machine learning model or a natural language processing model.

[1284] "Product candidate list" refers to data that lists multiple candidate products generated by a generative AI model based on user condition data.

[1285] "Formatting" refers to the process of converting the generated product candidate list into a format that is easy for users to view and understand.

[1286] "Receiving feedback" refers to the process of receiving additional conditions, opinions, etc. from users via input devices.

[1287] "Condition Modification" refers to the process of changing or modifying the original condition data based on feedback received from users.

[1288] "Invoking the generative AI model means again" refers to the process of generating a new product candidate list using the generative AI model again based on the modified condition data.

[1289] "Smartphone-based user interface means" refers to an interface that allows a user to interact with the system via a smartphone, input conditions, and send feedback.

[1290] "Means for providing feedback on additional conditions" refers to a process or interface that allows users to input additional requirements or requested changes and send them to the server.

[1291] System Overview

[1292] The present invention provides a system that allows users to select optimal products, particularly through a smartphone-based user interface, enabling them to find products efficiently and quickly. This system inputs user criteria data and generates an optimal product candidate list using a generative AI model based on that data. The system then presents the generated product candidate list to the user, receives feedback, and uses the generative AI model again to provide the optimal products that meet the user's needs.

[1293] Hardware and Software

[1294] The user's smartphone acts as the input interface, and the server receives the data, runs the generative AI model, generates a list of product candidates, and processes the feedback. For generative AI models, OpenAI GPT-4 and Google BERT are used as examples.

[1295] Program processing

[1296] 1. Enter your user data:

[1297] Users use their smartphones to access the chatbot screen on a dedicated application.

[1298] Users can input the necessary criteria for product selection via text or voice, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[1299] 2. Sending and receiving condition data:

[1300] The smartphone application converts the condition data entered by the user into JSON format and sends it to the server.

[1301] The server analyzes the received condition data and inputs it into the generative AI model.

[1302] 3. Run the generative AI model:

[1303] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[1304] The generative AI model searches and analyzes relevant product information from an internal database to extract the most suitable candidates.

[1305] 4. Generate and return product candidate list:

[1306] The server formats the generated product candidate list and converts it into a format that is easy for the user to understand.

[1307] The formatted product candidate list is sent back to the smartphone application and displayed on the user's screen.

[1308] Specific examples

[1309] If a user types "I'm looking for a medium-firm mattress under 50,000 yen" into a smartphone app, the system will operate as follows:

[1310] 1. A user types into a smartphone app, "I'm looking for a medium-firm mattress under 50,000 yen."

[1311] 2. This text data is converted to JSON format and sent to the server.

[1312] 3. The server inputs the data into a generative AI model (e.g., OpenAI GPT-4 or Google BERT) to generate a list of product candidates that match the criteria.

[1313] 4. Format the generated product candidate list and return it to the user in the following format:

[1314] "Mattress A, Price: 48,000 yen, Firmness: Medium, Size: Single, Brand: Brand X"

[1315] "Mattress B, Price: 45,000 yen, Firmness: Medium, Size: Single, Brand: Brand Y"

[1316] Prompt Sentence Examples

[1317] An example of a prompt sentence to input to a generative AI model is as follows.

[1318] "I'm looking for a mattress. My budget is under ¥50,000, I want a medium-firm mattress, and I don't have a specific brand in mind. Can you recommend the best product for me?"

[1319] In this way, it is possible to realize a system that allows users to quickly and efficiently select the most suitable product.

[1320] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1321] Step 1:

[1322] The user accesses the chatbot screen of the dedicated application using a smartphone. The user inputs the necessary product selection criteria by text or voice. For example, "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1323] Input: The condition data entered by the user.

[1324] Output: Condition data collected on the user's device.

[1325] Step 2:

[1326] The terminal converts the condition data entered by the user into JSON format and sends it to the server.

[1327] Input: Condition data.

[1328] Data processing: Conversion to JSON format.

[1329] Output: Criteria data converted to JSON format.

[1330] Step 3:

[1331] The server analyzes the received JSON formatted condition data and inputs it into the generative AI model.

[1332] Input: Criteria data in JSON format.

[1333] Data analysis: Parsing JSON data and applying it to generative AI models.

[1334] Output: Analysis results and inputs to generative AI models.

[1335] Step 4:

[1336] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[1337] Input: Parsed condition data.

[1338] Data computation: Generative AI models generate product candidate lists.

[1339] Output: Product candidate list.

[1340] Step 5:

[1341] The generated product candidate list is formatted on the server and converted into a format that is easy for users to understand.

[1342] Input: Product picklist.

[1343] Data processing: Formatting the product candidate list.

[1344] Output: A formatted list of product suggestions.

[1345] Step 6:

[1346] The server returns the formatted product candidate list to the user's smartphone application, which displays it on the user's screen.

[1347] Input: A formatted list of product suggestions.

[1348] Output: Product candidate list displayed on user's device.

[1349] Step 7:

[1350] The user checks the product candidate list, sets additional requirements and feedback, and re-enters the list. An example of feedback is, "Mattress A is good, but I want something a little cheaper."

[1351] Input: User feedback.

[1352] Output: Feedback data with added and modified conditions.

[1353] Step 8:

[1354] The server again receives feedback from the user and generates new JSON data that reflects the modified conditions.

[1355] Input: Feedback data.

[1356] Data processing: Generate JSON data that reflects the condition modifications.

[1357] Output: The modified JSON data.

[1358] Step 9:

[1359] Based on the revised condition data, the server calls the generative AI model again and generates a new list of product candidates.

[1360] Input: The modified condition data.

[1361] Data computation: Generative AI models generate new product candidate lists.

[1362] Output: A new list of product candidates.

[1363] Step 10:

[1364] The server then formats the regenerated product candidate list, converts it into a format that is easy for the user to understand, and returns it to the user's smartphone application.

[1365] Input: A new product shortlist.

[1366] Data processing: Formatting a new product candidate list.

[1367] Output: The formatted new product candidate list displayed on the user's device.

[1368] This process creates a system that allows users to efficiently select the most suitable product.

[1369] 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.

[1370] The present invention provides a system that reduces the time and effort required for users to select optimal products, especially mattresses. A key feature of the system is the incorporation of an emotion engine that recognizes the user's emotions and selects products that reflect their emotional state, enabling more personalized recommendations. Specific embodiments of the system are described below.

[1371] Overall system overview

[1372] In this system, users input criteria data via a chat-style interface, and the server then uses a generative AI model and an emotion engine to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[1373] User Action

[1374] Enter the condition:

[1375] Users access a chat-style input screen using their device and input their desired conditions (budget, firmness, size, brand, etc.) For example, they could input "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1376] Server Processing

[1377] Receiving user data:

[1378] The server receives the condition data sent by the user. The received data has the following format:

[1379] json

[1380] {

[1381] "budget": "50000",

[1382] "hardness": "medium",

[1383] "Size": "Single",

[1384] "Brand": "Not specified"

[1385] }

[1386] Calling the Emotion Engine:

[1387] Based on the received condition data, the server calls the emotion engine and analyzes the user's emotional state. The emotion engine recognizes emotions through chat context, user input, facial recognition, and voice analysis.

[1388] Processing generative AI models

[1389] Invoke the generative AI model:

[1390] The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[1391] Generate product picklist:

[1392] The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria and emotional state. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[1393] Server Processing

[1394] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[1395] json

[1396] [

[1397] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1398] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1399] ]

[1400] Sending results to the user:

[1401] The server returns the formatted product candidate list to the user's chat screen.

[1402] Feedback Processing

[1403] User feedback:

[1404] The user can review the provided list of product candidates and provide feedback on what is closest to their desired product or on any additional requirements. For example, they can enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[1405] Re-invoke the generated AI model:

[1406] The server receives feedback from the user, reflects the modified conditions and emotional state, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[1407] Regenerate product pick list

[1408] Generative AI models:

[1409] The generative AI model then analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[1410] server:

[1411] The server then formats the new product list and sends it back to the user, for example:

[1412] json

[1413] [

[1414] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[1415] ]

[1416] Final selection and purchase

[1417] Final decision:

[1418] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[1419] Purchase procedure:

[1420] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details of the process.

[1421] Specific examples

[1422] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1423] 2. The server receives the condition data, and the emotion engine recognizes the user's emotional state. At this time, it determines that the user is feeling stressed.

[1424] 3. The generative AI model uses big data to analyze the conditions and emotional state and suggests mattresses with relaxing effects.

[1425] 4. The server returns the generated candidate lists "Mattress A" and "Mattress B" to the user.

[1426] 5. A user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[1427] 6. The server passes the feedback and emotional state back to the generative AI model and suggests a new candidate, "Mattress C."

[1428] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[1429] The system allows users to find the perfect mattress in an efficient and emotionally sensitive manner.

[1430] The processing flow will be explained below.

[1431] Step 1:

[1432] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[1433] Step 2:

[1434] Server: The server receives the condition data sent by the user. The received data is in the following format:

[1435] json

[1436] {

[1437] "budget": "50000",

[1438] "hardness": "medium",

[1439] "Size": "Single",

[1440] "Brand": "Not specified"

[1441] }

[1442] Step 3:

[1443] Server: Based on the received condition data, the server calls the emotion engine to analyze the user's emotional state. The emotion engine recognizes the user's emotions using the user's input and other emotional indicators (e.g., context, facial recognition, voice analysis, etc.).

[1444] Step 4:

[1445] Emotion engine: The emotion engine recognizes the user's emotional state, for example, determining that the user is feeling stressed. The recognized emotional state is returned to the server.

[1446] Step 5:

[1447] Server: The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[1448] Step 6:

[1449] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the user's criteria and emotional state. This process involves running database queries, filtering, and finally ranking the products by scoring.

[1450] Step 7:

[1451] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be returned to the user. For example, the following candidate list is generated:

[1452] json

[1453] [

[1454] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1455] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1456] ]

[1457] Step 8:

[1458] Server: The server returns the formatted product candidate list to the user's chat screen.

[1459] Step 9:

[1460] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[1461] Step 10:

[1462] Server: The server receives feedback from the user and passes it back to the generative AI model along with condition modifications and emotional states.

[1463] Step 11:

[1464] Generative AI model: The generative AI model analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[1465] Step 12:

[1466] Server: The server formats the new product list and sends it back to the user, for example:

[1467] json

[1468] [

[1469] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[1470] ]

[1471] Step 13:

[1472] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[1473] Step 14:

[1474] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[1475] Example 2

[1476] 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."

[1477] Conventional product recommendation systems only suggest products based on the user's criteria, without taking into account the user's emotional state. This creates a problem in that they provide products without understanding the user's feelings when searching for a product. As a result, the suggested products often do not match the user's actual needs, making it difficult to increase satisfaction.

[1478] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an input device means for inputting user condition data, a server device means for receiving the user condition data, a generation AI model means for generating an optimal product candidate list based on the received condition data and the user's emotional state recognized by the emotion analysis means, a server device means for shaping the generated product candidate list and returning it to the user, and a server device means for receiving feedback from the user, reflecting the condition modifications and emotional state, and again calling the generation AI model means. This enables personalized product recommendations that take the user's emotional state into consideration.

[1479] "User" means a person who uses the System to select and purchase a Product.

[1480] "Condition data" is information about desired conditions (budget, hardness, size, brand, etc.) that a user enters into the system.

[1481] The "input device means" is a device (such as a personal computer or smartphone) that the user uses to input condition data.

[1482] "Server device means" refers to a server device for receiving and processing data from a user.

[1483] "Emotion analysis means" refers to technology that recognizes the user's emotional state through user input, facial recognition, voice analysis, etc.

[1484] The "generative AI model means" is an artificial intelligence model that analyzes big data based on the user's condition data and emotional state, and generates an optimal product candidate list.

[1485] A "product candidate list" is a plurality of candidate products selected by a generative AI model means and suggested to a user.

[1486] "Feedback" refers to the act of a user providing more desirable or additional conditions for a product candidate list.

[1487] "Condition modification" is the process of changing and updating the initial conditions based on user feedback.

[1488] "Invoking the generating AI model means again" is a procedure for causing the generating AI model means to generate a product candidate list again based on the condition modification and emotional state.

[1489] This invention relates to a system that reduces the time and effort required for users to select the optimal product, especially a mattress. This system is capable of providing more personalized recommendations by incorporating an emotion engine that recognizes the user's emotions and selects products that reflect the user's emotional state.

[1490] Overall system overview

[1491] In this system, users input criteria data through a chat-style interface, and the server generates and provides an optimal product candidate list using a generative AI model and emotion engine based on that data. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[1492] Specific Embodiments

[1493] User input operations

[1494] The user accesses the chat screen of the system using a terminal and inputs the desired data. For example, the user can input specific desired data such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference." This input information is input in text format as follows:

[1495] Budget: 50,000 yen or less, medium hardness, single size, no particular brand preference.

[1496] Server reception and analysis

[1497] The server receives the condition data sent by the user and analyzes it. The data is sent to the server in JSON format, and then the emotion engine is called to analyze the user's emotional state. The emotion engine uses the device's camera and microphone to perform facial recognition and voice analysis, and obtains an analysis result such as "the user is feeling stressed."

[1498] Invoking a generative AI model

[1499] The server passes the analysis results and condition data to the AI ​​model to generate the optimal product candidate list. The input prompt is set as follows:

[1500] Budget: ¥50,000 or less, medium hardness, single size, no particular brand. Users are feeling stressed.

[1501] The generative AI model analyzes large amounts of data and selects the best product candidates based on criteria and emotional state.

[1502] Generate and provide product candidate lists

[1503] The product candidate list generated by the generative AI model is formatted on the server and provided to the user. Specifically, the server replies to the user's chat screen with a message like the following:

[1504] The recommended mattresses are as follows:

[1505] 1. Mattress A - Price: 48,000 yen - Firmness: Medium - Size: Single - Brand X

[1506] 2. Mattress B - Price: 45,000 yen - Firmness: Medium - Size: Single - Brand Y

[1507] Receiving feedback and reanalyzing

[1508] The user reviews the provided product candidate list and provides feedback with more detailed requirements. For example, they may send feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback and calls the generative AI model again to begin the process of generating a new product candidate list.

[1509] Providing a regenerated candidate list

[1510] The server again formats the new product candidate list obtained from the generative AI model and provides it to the user in the following format:

[1511] This is a new proposal.

[1512] 1. Mattress C - Price: 42,000 yen - Firmness: Medium - Size: Single - Brand: Z

[1513] Final selection and checkout

[1514] The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C." The server receives the selection and guides the user to detailed information about the purchase procedure and a link to the purchase page.

[1515] This system allows users to efficiently select the most suitable product while taking into account their own emotional state.

[1516] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1517] Step 1: User Input

[1518] Users access the chat screen of the system using a terminal and input their desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1519] Input: User's desired conditions (budget, hardness, size, brand, etc.)

[1520] Output: The user's desired conditions are sent from the terminal in text format.

[1521] Step 2: Server Receive

[1522] The server receives the condition data sent by the user and prepares to analyze the received data.

[1523] Input: The user's desired conditions are sent to the server in text format.

[1524] Output: The criteria data is converted to JSON format and passed to the next processing step.

[1525] Step 3: Sentiment Analysis

[1526] The server calls the emotion engine based on the condition data and analyzes the user's emotional state. The emotion engine uses the device's camera and microphone to recognize the user's face and analyze their voice.

[1527] Input: Condition data (JSON format) and user input

[1528] Output: Sentiment analysis result (e.g., user is stressed)

[1529] Step 4: Call the generative AI model

[1530] The server passes the analyzed emotional state and condition data to a generative AI model, which then generates an optimal product candidate list. The generative AI model performs analysis using big data.

[1531] Input: Condition data and emotion analysis results

[1532] Output: Product candidate list generated by the generative AI model

[1533] Step 5: Product candidate list formatting

[1534] The server formats the product candidate list obtained from the generative AI model and prepares it for return to the user.

[1535] Input: Product candidate list generated by the generative AI model

[1536] Output: A formatted list of product candidates

[1537] Step 6: Reply to the user

[1538] The server sends the formatted product candidate list to the user's chat screen.

[1539] Input: A formatted list of product candidates

[1540] Output: Product candidate list displayed on the user's chat screen

[1541] Step 7: User feedback

[1542] Users can review the provided product shortlist and enter additional preferences and feedback, such as "Mattress A looks good, but let me know if there's anything cheaper."

[1543] Input: User additional requirements and feedback

[1544] Output: User feedback sent from the terminal in text format

[1545] Step 8: Server reception and reparsing

[1546] The server receives feedback sent by the user and invokes a regenerative AI model that reflects condition modifications and emotional states.

[1547] Input: User feedback (text format)

[1548] Output: Reanalyzed condition data and sentiment analysis results

[1549] Step 9: Regenerate AI model call

[1550] The server then calls the generative AI model again based on the reanalyzed condition data and emotional state, and generates a new list of product candidates.

[1551] Input: Reanalyzed condition data and emotional state

[1552] Output: A new product candidate list generated by the generative AI model

[1553] Step 10: Provide a regenerated candidate list

[1554] The server formats the regenerated product candidate list and provides it to the user's chat screen again.

[1555] Input: A new product candidate list generated by the regenerative AI model

[1556] Output: New, re-formatted product candidate list

[1557] Step 11: Final user selection

[1558] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[1559] Input: User's final selection

[1560] Output: The final selected product information is sent to the server.

[1561] Step 12: Purchase procedure guide

[1562] The server receives the final selected product information and provides the user with detailed purchase procedure information and a link to the purchase page.

[1563] Input: User's last selected product information

[1564] Output: A purchase procedure message will be displayed on the user's chat screen.

[1565] By combining the above steps, users can efficiently select the most suitable mattress and proceed with the purchase process while taking into account their own emotional state.

[1566] (Application example 2)

[1567] 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."

[1568] Conventional product selection systems suggest products based on user condition data, but because they do not take the user's emotional state into account, they have the problem of not being able to make suggestions that are optimal for the user's psychological state. In addition, there is no way to display product location information within the store, making it difficult for users to search for products efficiently. This reduces users' motivation to purchase, and there are issues such as the time and effort required to select the most suitable product.

[1569] 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: an input means for inputting user condition data; a communication means for receiving the user condition data; a generation AI model means for analyzing the received condition data and emotion analysis data to generate an optimal product candidate list; a communication means for formatting the generated product candidate list and returning it to the user; a communication means for reflecting condition modifications based on user feedback and emotion analysis data and again calling the generation AI model means; an emotion recognition means for recognizing the user's emotional state in real time; and a visualization means for displaying product locations within a physical store for the user. This enables personalized product suggestions based on the user's emotional state, allowing the user to efficiently search for products within the store.

[1570] "User condition data" refers to the desired conditions (budget, hardness, size, brand, etc.) entered by the user.

[1571] "Input means" refers to a device or interface that allows a user to input condition data.

[1572] "Communication means" refers to a network interface for sending and receiving data.

[1573] "Emotion analysis data" is data that represents the user's emotional state after analyzing it.

[1574] "Generative AI model means" refers to an artificial intelligence model for generating an optimal product candidate list using received data.

[1575] "Emotion recognition means" refers to technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state in real time.

[1576] "Visualization means" refers to displays and AR technologies that allow users to visually view information.

[1577] "Product Candidate List" means a list of optimal products generated by the generative AI model means and based on the user's criteria and emotional state.

[1578] "Feedback" refers to any additional preferences or comments provided by the User.

[1579] "Physical store" refers to an actual store space.

[1580] To implement this invention, a system including the following components is required. This system is designed to suggest optimal products based on a user's condition data and emotion data. Specifically, it uses a user input means, a communication means, an emotion recognition means, a generation AI model means, and a visualization means.

[1581] First, the user puts on the smart glasses and inputs their requirements by voice or touch. This includes information such as budget, hardness, size, and brand. For example, "budget is under ¥50,000, hardness is medium, size is single, and brand is not specified."

[1582] The input condition data is sent to the server via a communication means. The server receives the data and analyzes the user's emotional state via an emotion recognition means. Emotion recognition uses the smart glasses' camera and microphone to analyze the user's facial expressions and tone of voice. For example, it can determine whether the user is feeling stressed.

[1583] The server then passes the received condition data and emotion data to the generative AI model means, which analyzes this information using big data. The generative AI model generates an optimal product candidate list based on the condition and emotion state. The generated list is sent back to the smart glasses via the communication means.

[1584] The smart glasses use a visualization tool to present the generated product candidate list to the user, along with the product location information, allowing the user to efficiently search for products in the store.

[1585] For example, if "Mattress A" is included in the generated candidate list, its location information will be displayed on the display. The user can walk around the store while checking this information and check the products directly. If the user gives feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," this feedback will also be sent to the server via the communication means.

[1586] The server then calls the generative AI model again based on the received feedback and emotional data, updating the product candidate list. In this way, the user's feedback and emotional state are reflected in real time, allowing for smooth selection of the most suitable product.

[1587] As a concrete example, the following prompt sentence is input to the generative AI model:

[1588] "List products (mattresses) that match the following criteria: Conditions: { "Budget": "50000", "Firmness": "Medium", "Size": "Single", "Brand": "Not Specified"} User emotional state: "Feeling stressed""

[1589] The system allows users to find the perfect mattress in an efficient and emotionally sensitive way.

[1590] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1591] Step 1:

[1592] The user puts on the smart glasses and inputs the desired conditions by voice or touch operation. The input conditions include budget, hardness, size, brand, etc. Data such as "budget within 50,000 yen, hardness medium, size single, no particular brand" is input. The input data is temporarily stored in the smart glasses' internal memory.

[1593] Step 2:

[1594] The user's input data is sent to the server via a communication method, either via the internet or the store's Wi-Fi. The server processes the received condition data as data for analysis. At this time, the received data is converted to JSON format and saved in the server's data storage.

[1595] Step 3:

[1596] The server then calls the emotion recognition unit based on the received data. It uses the smart glasses' camera and microphone to collect the user's facial expressions and tone of voice, analyzes them, and recognizes their emotional state. The analyzed emotional state is then classified into categories such as "feeling stressed" and output as data.

[1597] Step 4:

[1598] The server passes the condition data and the emotion data obtained by emotion recognition to the generative AI model means. The generative AI model receives these data as input and performs analysis using big data. The analysis process includes running database queries, filtering, and scoring. The generative AI model outputs a list of optimal product candidates, which is sent back to the server in JSON format.

[1599] Step 5:

[1600] The server then formats the received product candidate list and converts it into a data format for sending back to the user. This conversion process includes formatting data such as product name, price, and location information. The formatted data is then sent back to the smart glasses via a communication means.

[1601] Step 6:

[1602] The smart glasses display the received product candidate list to the user using a visualization means. The smart glasses display shows the names, prices, and location information of the candidate products in the store. The user confirms the displayed information and moves around the store to actually check the products.

[1603] Step 7:

[1604] After checking the products, the user can provide feedback by voice input or touch operation on the smart glasses. For example, they can input feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." This feedback is sent to the server via a communication means.

[1605] Step 8:

[1606] The server receives the feedback and updates the condition data and emotion data. Based on the updated data, it calls the generation AI model means again to generate a new product candidate list. This new candidate list is also shaped by the above process and sent to the smart glasses again.

[1607] Step 9:

[1608] The smart glasses will then display the new product candidate list using a visualization tool and present it to the user. The user will then review the products based on the information provided and make a final selection. After making their selection, they will be guided through the purchase process.

[1609] Through these steps, users can efficiently select the best mattress for them. This system provides personalized product recommendations based on the user's emotional state and efficient in-store navigation.

[1610] 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.

[1611] 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.

[1612] 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.

[1613] [Fourth embodiment]

[1614] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1615] 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.

[1616] 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).

[1617] 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.

[1618] 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.

[1619] 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).

[1620] 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.

[1621] 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.

[1622] 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.

[1623] 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.

[1624] 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.

[1625] 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.

[1626] 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."

[1627] The present invention provides a system that reduces the time and effort required for a user to select an optimal product, particularly a mattress. Specific embodiments of the system are described below.

[1628] Overall system overview

[1629] In this system, users input criteria data via a chat-style interface, and the server uses a generative AI model to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list using the generative AI model again based on that feedback.

[1630] User Action

[1631] Enter the condition:

[1632] Users access a chat-style input screen using their device and enter their desired conditions, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[1633] Server Processing

[1634] Receiving user data:

[1635] The server receives the condition data sent by the user. The received data has the following format:

[1636] json

[1637] {

[1638] "budget": "50000",

[1639] "hardness": "medium",

[1640] "Size": "Single",

[1641] "Brand": "Not specified"

[1642] }

[1643] Invoke the generative AI model:

[1644] The server passes the received condition data to the generative AI model, which analyzes the big data. The generative AI model retrieves relevant mattress data from the database and generates a list of optimal product candidates that match the conditions.

[1645] Generate candidate list:

[1646] Based on the information obtained from the AI ​​model, the server formats the product candidate list and converts it into a format that can be sent back to the user. For example, the following candidate list is generated:

[1647] json

[1648] [

[1649] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1650] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1651] ]

[1652] Sending results to the user:

[1653] The server returns the formatted candidate list to the user's chat screen.

[1654] Feedback Processing

[1655] User feedback:

[1656] The user can review the provided list of candidates and provide feedback on what is closest to their desired fit or on any additional requirements. For example, they could enter, "Mattress A seems good, but let me know if there is anything cheaper."

[1657] Re-invoke the generated AI model:

[1658] The server receives feedback from the user, reflects the revised conditions, and invokes the generative AI model again. This process generates the optimal candidate list based on the new conditions from the database, just as it did last time.

[1659] Generate a new candidate list:

[1660] Based on the revised conditions, the server again reformats the information obtained from the generative AI model and returns a new product candidate list to the user. For example, the following additional candidates are generated:

[1661] json

[1662] [

[1663] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[1664] ]

[1665] Final selection and purchase

[1666] Final decision:

[1667] The user finally selects the best product from the presented options and makes a decision, such as "I will purchase Mattress C."

[1668] Purchase procedure:

[1669] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[1670] Specific examples

[1671] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1672] 2. The server receives the condition data, and the generative AI model uses big data to generate the optimal product candidate list.

[1673] 3. The server returns the generated candidate list to the user, presenting "Mattress A" and "Mattress B."

[1674] 4. The user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[1675] 5. The server invokes the generative AI model again based on the feedback and generates a new candidate list.

[1676] 6. The server returns a new candidate, "Mattress C," to the user.

[1677] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[1678] In this way, the system allows users to efficiently find the best mattress.

[1679] The processing flow will be explained below.

[1680] Step 1:

[1681] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[1682] Step 2:

[1683] Server: The server receives the condition data sent by the user. The received data is in the following format:

[1684] json

[1685] {

[1686] "budget": "50000",

[1687] "hardness": "medium",

[1688] "Size": "Single",

[1689] "Brand": "Not specified"

[1690] }

[1691] Step 3:

[1692] Server: The server passes the received condition data to the generative AI model and invokes it. The generative AI model analyzes the condition data using big data and generates an optimal product candidate list.

[1693] Step 4:

[1694] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[1695] Step 5:

[1696] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[1697] json

[1698] [

[1699] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1700] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1701] ]

[1702] Step 6:

[1703] Server: The server returns the formatted product candidate list to the user's chat screen.

[1704] Step 7:

[1705] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[1706] Step 8:

[1707] Server: The server receives feedback from the user, reflects the revised conditions, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[1708] Step 9:

[1709] Generative AI model: The generative AI model analyzes the condition data again and generates a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[1710] Step 10:

[1711] Server: The server formats the new product list and sends it back to the user, for example:

[1712] json

[1713] [

[1714] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[1715] ]

[1716] Step 11:

[1717] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[1718] Step 12:

[1719] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[1720] Example 1

[1721] 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."

[1722] In the past, when users selected products such as mattresses, it took a lot of time and effort to find the best product from the vast number of options available. Furthermore, manually searching and comparing product information was cumbersome, making it difficult for users to find the product that best met their requirements.

[1723] 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.

[1724] In this invention, the server includes means for inputting user condition data, data processing means for receiving the user condition data, machine learning model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, data processing means for processing the generated product candidate list and returning it to the user, and data processing means for receiving feedback from the user, reflecting condition modifications, and again calling the machine learning model means, thereby enabling users to efficiently find products that best fit their conditions.

[1725] "User condition data" refers to the desired conditions that a user inputs to select a specific product, and includes data such as budget, hardness, size, brand, etc.

[1726] "Means for inputting user condition data" refers to devices or software that provide an interface for users to input their desired conditions. This includes smartphones, tablets, PC applications, and web browsers.

[1727] "Data processing means for receiving user condition data" refers to a server or network device for receiving and processing the condition data sent by the user.

[1728] "Big data" refers to a huge amount of data, a dataset that can be analyzed to extract valuable information, such as product information, reviews, and pricing data.

[1729] A "machine learning model" is an algorithm or program used to analyze data and generate an optimal list of product candidates based on the results. A specific example is a generative AI model.

[1730] The "generative machine learning model means" refers to software and hardware for receiving user condition data and large-scale data as input and executing a machine learning model to generate an optimal product candidate list.

[1731] "Data processing means" refers to the server and related software for receiving, processing, transmitting and feedback processing data.

[1732] "Product candidate list" refers to a list of product candidates generated based on the user's criteria and presented to the user for consideration.

[1733] "Feedback" refers to the user inputting their opinions and further requirements regarding the provided product candidate list.

[1734] "Condition modification" refers to changing or adding to the original condition data based on user feedback.

[1735] "Invoke" refers to launching a specific function or process (in this case, a machine learning model) to execute it.

[1736] The present invention provides a system for efficiently selecting a desired product, particularly a mattress, by a user. Specific embodiments will be described below.

[1737] Hardware and Software Use

[1738] User condition data input

[1739] The user accesses a chat-style input screen using a web browser or dedicated application on a device (smartphone, tablet, PC, etc.) and enters desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1740] Receiving and processing condition data

[1741] The server receives the condition data sent by the user as an HTTP POST request and parses it in JSON format. The data can be processed using a web server such as NGINX or Apache, or a programming language such as Python or Node.js. The received data is in the following format:

[1742] json

[1743] {

[1744] "budget": "50000",

[1745] "hardness": "medium",

[1746] "Size": "Single",

[1747] "Brand": "Not specified"

[1748] }

[1749] Invoking a generative AI model

[1750] The server invokes a generative AI model (e.g., OpenAI's GPT-3 or Google's BERT) based on the received condition data. The prompt is created as follows:

[1751] "Choose the best mattress for you, with a budget of under ¥50,000, medium firmness, single size, and no particular brand preference."

[1752] Generate product candidate list

[1753] The generative AI model analyzes large-scale data and generates a list of product candidates that match the criteria. The generated list of product candidates is received by the server and formatted to be sent back to the user. Examples of formatting are as follows:

[1754] "Mattress A: Price 48,000 yen, medium firmness, size single, brand X"

[1755] "Mattress B: Price 45,000 yen, medium firmness, size single, brand Y"

[1756] Sending results to the user

[1757] The server returns the formatted product candidate list to the user's chat screen, and the user confirms the provided list.

[1758] Feedback Processing

[1759] The user reviews the provided candidate list and provides feedback on what is closest to their desired outcome or on additional requirements. For example, they might input, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback, reflects the revised requirements, and invokes the generative AI model again. The new prompt text becomes:

[1760] "Choose the best mattress that is medium firm, single size, and doesn't specify the brand, with a budget of 45,000 yen or less, if possible."

[1761] The newly generated product candidate list is formatted in the same way and sent back to the user.

[1762] Specific examples

[1763] When a user inputs "budget under ¥50,000, medium firmness, single size, no particular brand," the server receives the condition data and the generative AI model uses big data to generate an optimal product candidate list. On the other hand, if the user inputs feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," the server readjusts the conditions and sends a new prompt to the generative AI model, generating a new product candidate list.

[1764] In this way, the present invention is a system that efficiently generates optimal product candidates based on condition data, significantly reducing the time and effort required by the user.

[1765] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1766] Step 1:

[1767] The user accesses a chat-style input screen using a device (smartphone, tablet, PC, etc.) and enters the desired conditions. For example, they might enter "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1768] Input: User's desired conditions (budget, hardness, size, brand)

[1769] Output: User's desired conditions data

[1770] Step 2:

[1771] The device sends the user's criteria data to the server, which receives it as an HTTP POST request and parses it into JSON format.

[1772] Input: User's desired conditions data

[1773] Output: Condition data in JSON format

[1774] Step 3:

[1775] The server uses the received condition data to create a prompt for the generative AI model, such as "Please choose the best mattress for me, within a budget of ¥50,000, with medium firmness, single size, and no particular brand preference."

[1776] Input: Condition data in JSON format

[1777] Output: prompt statement

[1778] Step 4:

[1779] The server sends the prompt to a generative AI model (such as OpenAI's GPT-3 or BERT), which analyzes the prompt and generates an optimal product candidate list from large-scale data.

[1780] Input: prompt statement

[1781] Output: Raw product candidate list data

[1782] Step 5:

[1783] The server receives the raw data of the product candidate list returned by the generative AI model and formats it into a format that is easy for users to understand, such as "Mattress A: Price 48,000 yen, Medium firmness, Size single, Brand X."

[1784] Input: Raw data of product candidate list

[1785] Output: A formatted list of product candidates

[1786] Step 6:

[1787] The server returns the formatted product candidate list to the user's chat screen, which the user can confirm.

[1788] Input: A formatted list of product candidates

[1789] Output: A list of product candidates presented to the user

[1790] Step 7:

[1791] The user enters feedback on the presented product candidate list. For example, "Mattress A seems good, but please let me know if there is anything cheaper."

[1792] Input: User feedback

[1793] Output: Feedback data

[1794] Step 8:

[1795] The device sends the user's feedback data to the server. The server receives this data and modifies the condition data. It then creates a new prompt and sends it back to the generative AI model. For example, it generates a prompt such as, "Please choose the best mattress for me, preferably within 45,000 yen, with medium firmness, single size, and no particular brand preference."

[1796] Input: Feedback data

[1797] Output: Modified prompt statement

[1798] Step 9:

[1799] The generative AI model generates a new product candidate list based on the new prompt. The server receives this and formats it again. The formatted list is then sent to the user again.

[1800] Input: modified prompt text

[1801] Output: Regenerated product candidate list

[1802] Step 10:

[1803] Finally, the user selects the most suitable product from the presented product candidates and notifies the system of their intention to purchase the selected product. The server provides the user with information regarding the purchase procedure.

[1804] Input: Final selected product data

[1805] Output: Purchase procedure guide information

[1806] (Application example 1)

[1807] 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."

[1808] Current online shopping sites have the problem that it is difficult for users to efficiently find the products they want. In particular, because they do not provide prompt responses to input of conditions or feedback, users must spend a great deal of time and effort to find the optimal product. The purpose of this invention is to solve these problems and provide a system that allows users to efficiently select the optimal product.

[1809] 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.

[1810] In this invention, the server includes means for inputting user condition data, means for receiving the user condition data, a generation AI model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list, means for formatting the generated product candidate list and returning it to the user, means for receiving feedback from the user and reflecting the condition modifications to call the generation AI model means again, a user interface means using a smartphone, and means for presenting the generated product candidate list to the user and providing feedback on additional conditions, thereby enabling the user to quickly and efficiently select optimal products.

[1811] "User condition data" refers to the desired conditions entered by the user when selecting a product, such as budget, size, hardness, brand, and other specific requirements.

[1812] "Input means" refers to the interface or device through which the user inputs condition data, such as a smartphone, PC, or tablet.

[1813] "Server means" refers to a computer system for receiving, storing, and analyzing data sent by users, including databases and cloud servers.

[1814] "Generative AI model means" refers to an artificial intelligence model used to analyze received data and generate optimal product candidates, including, for example, a machine learning model or a natural language processing model.

[1815] "Product candidate list" refers to data that lists multiple candidate products generated by a generative AI model based on user condition data.

[1816] "Formatting" refers to the process of converting the generated product candidate list into a format that is easy for users to view and understand.

[1817] "Receiving feedback" refers to the process of receiving additional conditions, opinions, etc. from users via input devices.

[1818] "Condition Modification" refers to the process of changing or modifying the original condition data based on feedback received from users.

[1819] "Invoking the generative AI model means again" refers to the process of generating a new product candidate list using the generative AI model again based on the modified condition data.

[1820] "Smartphone-based user interface means" refers to an interface that allows a user to interact with the system via a smartphone, input conditions, and send feedback.

[1821] "Means for providing feedback on additional conditions" refers to a process or interface that allows users to input additional requirements or requested changes and send them to the server.

[1822] System Overview

[1823] The present invention provides a system that allows users to select optimal products, particularly through a smartphone-based user interface, enabling them to find products efficiently and quickly. This system inputs user criteria data and generates an optimal product candidate list using a generative AI model based on that data. The system then presents the generated product candidate list to the user, receives feedback, and uses the generative AI model again to provide the optimal products that meet the user's needs.

[1824] Hardware and Software

[1825] The user's smartphone acts as the input interface, and the server receives the data, runs the generative AI model, generates a list of product candidates, and processes the feedback. For generative AI models, OpenAI GPT-4 and Google BERT are used as examples.

[1826] Program processing

[1827] 1. Enter your user data:

[1828] Users use their smartphones to access the chatbot screen on a dedicated application.

[1829] Users can input the necessary criteria for product selection via text or voice, such as "budget under 50,000 yen, medium firmness, single size, no particular brand preference."

[1830] 2. Sending and receiving condition data:

[1831] The smartphone application converts the condition data entered by the user into JSON format and sends it to the server.

[1832] The server analyzes the received condition data and inputs it into the generative AI model.

[1833] 3. Run the generative AI model:

[1834] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[1835] The generative AI model searches and analyzes relevant product information from an internal database to extract the most suitable candidates.

[1836] 4. Generate and return product candidate list:

[1837] The server formats the generated product candidate list and converts it into a format that is easy for the user to understand.

[1838] The formatted product candidate list is sent back to the smartphone application and displayed on the user's screen.

[1839] Specific examples

[1840] If a user types "I'm looking for a medium-firm mattress under 50,000 yen" into a smartphone app, the system will operate as follows:

[1841] 1. A user types into a smartphone app, "I'm looking for a medium-firm mattress under 50,000 yen."

[1842] 2. This text data is converted to JSON format and sent to the server.

[1843] 3. The server inputs the data into a generative AI model (e.g., OpenAI GPT-4 or Google BERT) to generate a list of product candidates that match the criteria.

[1844] 4. Format the generated product candidate list and return it to the user in the following format:

[1845] "Mattress A, Price: 48,000 yen, Firmness: Medium, Size: Single, Brand: Brand X"

[1846] "Mattress B, Price: 45,000 yen, Firmness: Medium, Size: Single, Brand: Brand Y"

[1847] Prompt Sentence Examples

[1848] An example of a prompt sentence to input to a generative AI model is as follows.

[1849] "I'm looking for a mattress. My budget is under ¥50,000, I want a medium-firm mattress, and I don't have a specific brand in mind. Can you recommend the best product for me?"

[1850] In this way, it is possible to realize a system that allows users to quickly and efficiently select the most suitable product.

[1851] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1852] Step 1:

[1853] The user accesses the chatbot screen of the dedicated application using a smartphone. The user inputs the necessary product selection criteria by text or voice. For example, "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1854] Input: The condition data entered by the user.

[1855] Output: Condition data collected on the user's device.

[1856] Step 2:

[1857] The terminal converts the condition data entered by the user into JSON format and sends it to the server.

[1858] Input: Condition data.

[1859] Data processing: Conversion to JSON format.

[1860] Output: Criteria data converted to JSON format.

[1861] Step 3:

[1862] The server analyzes the received JSON formatted condition data and inputs it into the generative AI model.

[1863] Input: Criteria data in JSON format.

[1864] Data analysis: Parsing JSON data and applying it to generative AI models.

[1865] Output: Analysis results and inputs to generative AI models.

[1866] Step 4:

[1867] The server uses a generative AI model (e.g., OpenAI GPT-4) to generate an optimal product candidate list based on the received condition data.

[1868] Input: Parsed condition data.

[1869] Data computation: Generative AI models generate product candidate lists.

[1870] Output: Product candidate list.

[1871] Step 5:

[1872] The generated product candidate list is formatted on the server and converted into a format that is easy for users to understand.

[1873] Input: Product picklist.

[1874] Data processing: Formatting the product candidate list.

[1875] Output: A formatted list of product suggestions.

[1876] Step 6:

[1877] The server returns the formatted product candidate list to the user's smartphone application, which displays it on the user's screen.

[1878] Input: A formatted list of product suggestions.

[1879] Output: Product candidate list displayed on user's device.

[1880] Step 7:

[1881] The user checks the product candidate list, sets additional requirements and feedback, and re-enters the list. An example of feedback is, "Mattress A is good, but I want something a little cheaper."

[1882] Input: User feedback.

[1883] Output: Feedback data with added and modified conditions.

[1884] Step 8:

[1885] The server again receives feedback from the user and generates new JSON data that reflects the modified conditions.

[1886] Input: Feedback data.

[1887] Data processing: Generate JSON data that reflects the condition modifications.

[1888] Output: The modified JSON data.

[1889] Step 9:

[1890] Based on the revised condition data, the server calls the generative AI model again and generates a new list of product candidates.

[1891] Input: The modified condition data.

[1892] Data computation: Generative AI models generate new product candidate lists.

[1893] Output: A new list of product candidates.

[1894] Step 10:

[1895] The server then formats the regenerated product candidate list, converts it into a format that is easy for the user to understand, and returns it to the user's smartphone application.

[1896] Input: A new product shortlist.

[1897] Data processing: Formatting a new product candidate list.

[1898] Output: The formatted new product candidate list displayed on the user's device.

[1899] This process creates a system that allows users to efficiently select the most suitable product.

[1900] 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.

[1901] The present invention provides a system that reduces the time and effort required for users to select optimal products, especially mattresses. A key feature of the system is the incorporation of an emotion engine that recognizes the user's emotions and selects products that reflect their emotional state, enabling more personalized recommendations. Specific embodiments of the system are described below.

[1902] Overall system overview

[1903] In this system, users input criteria data via a chat-style interface, and the server then uses a generative AI model and an emotion engine to generate and provide an optimal product candidate list. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[1904] User Action

[1905] Enter the condition:

[1906] Users access a chat-style input screen using their device and input their desired conditions (budget, firmness, size, brand, etc.) For example, they could input "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1907] Server Processing

[1908] Receiving user data:

[1909] The server receives the condition data sent by the user. The received data has the following format:

[1910] json

[1911] {

[1912] "budget": "50000",

[1913] "hardness": "medium",

[1914] "Size": "Single",

[1915] "Brand": "Not specified"

[1916] }

[1917] Calling the Emotion Engine:

[1918] Based on the received condition data, the server calls the emotion engine and analyzes the user's emotional state. The emotion engine recognizes emotions through chat context, user input, facial recognition, and voice analysis.

[1919] Processing generative AI models

[1920] Invoke the generative AI model:

[1921] The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[1922] Generate product picklist:

[1923] The generative AI model retrieves relevant mattress data from the database and selects the best products that match the criteria and emotional state. This process begins with a database query, followed by filtering, and finally ranking by scoring.

[1924] Server Processing

[1925] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be sent back to the user, for example:

[1926] json

[1927] [

[1928] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1929] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1930] ]

[1931] Sending results to the user:

[1932] The server returns the formatted product candidate list to the user's chat screen.

[1933] Feedback Processing

[1934] User feedback:

[1935] The user can review the provided list of product candidates and provide feedback on what is closest to their desired product or on any additional requirements. For example, they can enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[1936] Re-invoke the generated AI model:

[1937] The server receives feedback from the user, reflects the modified conditions and emotional state, and calls the generative AI model again to generate a new product candidate list based on the new conditions.

[1938] Regenerate product pick list

[1939] Generative AI models:

[1940] The generative AI model then analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[1941] server:

[1942] The server then formats the new product list and sends it back to the user, for example:

[1943] json

[1944] [

[1945] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[1946] ]

[1947] Final selection and purchase

[1948] Final decision:

[1949] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[1950] Purchase procedure:

[1951] The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details of the process.

[1952] Specific examples

[1953] 1. The user enters "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[1954] 2. The server receives the condition data, and the emotion engine recognizes the user's emotional state. At this time, it determines that the user is feeling stressed.

[1955] 3. The generative AI model uses big data to analyze the conditions and emotional state and suggests mattresses with relaxing effects.

[1956] 4. The server returns the generated candidate lists "Mattress A" and "Mattress B" to the user.

[1957] 5. A user gives feedback saying, "Mattress A feels good, but can you tell me if there's anything cheaper?"

[1958] 6. The server passes the feedback and emotional state back to the generative AI model and suggests a new candidate, "Mattress C."

[1959] 7. The user decides to purchase Mattress C, and the server guides them through the purchase process.

[1960] The system allows users to find the perfect mattress in an efficient and emotionally sensitive manner.

[1961] The processing flow will be explained below.

[1962] Step 1:

[1963] User: The user accesses a chat-style input screen using a device and inputs their desired conditions (budget, firmness, size, brand, etc.). For example, they might input "budget within 50,000 yen, firmness medium, size single, no particular brand preference."

[1964] Step 2:

[1965] Server: The server receives the condition data sent by the user. The received data is in the following format:

[1966] json

[1967] {

[1968] "budget": "50000",

[1969] "hardness": "medium",

[1970] "Size": "Single",

[1971] "Brand": "Not specified"

[1972] }

[1973] Step 3:

[1974] Server: Based on the received condition data, the server calls the emotion engine to analyze the user's emotional state. The emotion engine recognizes the user's emotions using the user's input and other emotional indicators (e.g., context, facial recognition, voice analysis, etc.).

[1975] Step 4:

[1976] Emotion engine: The emotion engine recognizes the user's emotional state, for example, determining that the user is feeling stressed. The recognized emotional state is returned to the server.

[1977] Step 5:

[1978] Server: The server passes the user's condition data and the recognized emotional state to the generative AI model, which then analyzes this information using big data to generate an optimal product candidate list.

[1979] Step 6:

[1980] Generative AI Model: The generative AI model retrieves relevant mattress data from the database and selects the best products that match the user's criteria and emotional state. This process involves running database queries, filtering, and finally ranking the products by scoring.

[1981] Step 7:

[1982] Server: The server receives the product candidate list obtained from the generative AI model and formats it to be returned to the user. For example, the following candidate list is generated:

[1983] json

[1984] [

[1985] {"Name": "Mattress A", "Price": "48000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand X"},

[1986] {"Name": "Mattress B", "Price": "45000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Y"}

[1987] ]

[1988] Step 8:

[1989] Server: The server returns the formatted product candidate list to the user's chat screen.

[1990] Step 9:

[1991] User: The user reviews the provided product candidate list and provides feedback on what is closest to their desired product or additional requirements. For example, they might enter, "Mattress A looks good, but please let me know if there is anything cheaper."

[1992] Step 10:

[1993] Server: The server receives feedback from the user and passes it back to the generative AI model along with condition modifications and emotional states.

[1994] Step 11:

[1995] Generative AI model: The generative AI model analyzes the condition data and emotional state again to generate a new list of optimal product candidates. For example, it adds "Mattress C" as a new candidate.

[1996] Step 12:

[1997] Server: The server formats the new product list and sends it back to the user, for example:

[1998] json

[1999] [

[2000] {"Name": "Mattress C", "Price": "42000", "Firmness": "Medium", "Size": "Single", "Brand": "Brand Z"}

[2001] ]

[2002] Step 13:

[2003] User: The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C."

[2004] Step 14:

[2005] Server: The server receives the user's selection and provides information about the purchase process, including a link to the purchase page and details about the process.

[2006] Example 2

[2007] 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."

[2008] Conventional product recommendation systems only suggest products based on the user's criteria, without taking into account the user's emotional state. This creates a problem in that they provide products without understanding the user's feelings when searching for a product. As a result, the suggested products often do not match the user's actual needs, making it difficult to increase satisfaction.

[2009] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes an input device means for inputting user condition data, a server device means for receiving the user condition data, a generation AI model means for generating an optimal product candidate list based on the received condition data and the user's emotional state recognized by the emotion analysis means, a server device means for shaping the generated product candidate list and returning it to the user, and a server device means for receiving feedback from the user, reflecting the condition modifications and emotional state, and again calling the generation AI model means. This enables personalized product recommendations that take the user's emotional state into consideration.

[2010] "User" means a person who uses the System to select and purchase a Product.

[2011] "Condition data" is information about desired conditions (budget, hardness, size, brand, etc.) that a user enters into the system.

[2012] The "input device means" is a device (such as a personal computer or smartphone) that the user uses to input condition data.

[2013] "Server device means" refers to a server device for receiving and processing data from a user.

[2014] "Emotion analysis means" refers to technology that recognizes the user's emotional state through user input, facial recognition, voice analysis, etc.

[2015] The "generative AI model means" is an artificial intelligence model that analyzes big data based on the user's condition data and emotional state, and generates an optimal product candidate list.

[2016] A "product candidate list" is a plurality of candidate products selected by a generative AI model means and suggested to a user.

[2017] "Feedback" refers to the act of a user providing more desirable or additional conditions for a product candidate list.

[2018] "Condition modification" is the process of changing and updating the initial conditions based on user feedback.

[2019] "Invoking the generating AI model means again" is a procedure for causing the generating AI model means to generate a product candidate list again based on the condition modification and emotional state.

[2020] This invention relates to a system that reduces the time and effort required for users to select the optimal product, especially a mattress. This system is capable of providing more personalized recommendations by incorporating an emotion engine that recognizes the user's emotions and selects products that reflect the user's emotional state.

[2021] Overall system overview

[2022] In this system, users input criteria data through a chat-style interface, and the server generates and provides an optimal product candidate list using a generative AI model and emotion engine based on that data. Users can review the candidate list and provide feedback, and the server updates the candidate list again using the generative AI model based on that feedback and their emotional state.

[2023] Specific Embodiments

[2024] User input operations

[2025] The user accesses the chat screen of the system using a terminal and inputs the desired data. For example, the user can input specific desired data such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference." This input information is input in text format as follows:

[2026] Budget: 50,000 yen or less, medium hardness, single size, no particular brand preference.

[2027] Server reception and analysis

[2028] The server receives the condition data sent by the user and analyzes it. The data is sent to the server in JSON format, and then the emotion engine is called to analyze the user's emotional state. The emotion engine uses the device's camera and microphone to perform facial recognition and voice analysis, and obtains an analysis result such as "the user is feeling stressed."

[2029] Invoking a generative AI model

[2030] The server passes the analysis results and condition data to the AI ​​model to generate the optimal product candidate list. The input prompt is set as follows:

[2031] Budget: ¥50,000 or less, medium hardness, single size, no particular brand. Users are feeling stressed.

[2032] The generative AI model analyzes large amounts of data and selects the best product candidates based on criteria and emotional state.

[2033] Generate and provide product candidate lists

[2034] The product candidate list generated by the generative AI model is formatted on the server and provided to the user. Specifically, the server replies to the user's chat screen with a message like the following:

[2035] The recommended mattresses are as follows:

[2036] 1. Mattress A - Price: 48,000 yen - Firmness: Medium - Size: Single - Brand X

[2037] 2. Mattress B - Price: 45,000 yen - Firmness: Medium - Size: Single - Brand Y

[2038] Receiving feedback and reanalyzing

[2039] The user reviews the provided product candidate list and provides feedback with more detailed requirements. For example, they may send feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." The server receives this feedback and calls the generative AI model again to begin the process of generating a new product candidate list.

[2040] Providing a regenerated candidate list

[2041] The server again formats the new product candidate list obtained from the generative AI model and provides it to the user in the following format:

[2042] This is a new proposal.

[2043] 1. Mattress C - Price: 42,000 yen - Firmness: Medium - Size: Single - Brand: Z

[2044] Final selection and checkout

[2045] The user finally selects the best product from the presented product candidates and decides to "purchase Mattress C." The server receives the selection and guides the user to detailed information about the purchase procedure and a link to the purchase page.

[2046] This system allows users to efficiently select the most suitable product while taking into account their own emotional state.

[2047] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2048] Step 1: User Input

[2049] Users access the chat screen of the system using a terminal and input their desired conditions, such as "budget within 50,000 yen, medium firmness, single size, no particular brand preference."

[2050] Input: User's desired conditions (budget, hardness, size, brand, etc.)

[2051] Output: The user's desired conditions are sent from the terminal in text format.

[2052] Step 2: Server Receive

[2053] The server receives the condition data sent by the user and prepares to analyze the received data.

[2054] Input: The user's desired conditions are sent to the server in text format.

[2055] Output: The criteria data is converted to JSON format and passed to the next processing step.

[2056] Step 3: Sentiment Analysis

[2057] The server calls the emotion engine based on the condition data and analyzes the user's emotional state. The emotion engine uses the device's camera and microphone to recognize the user's face and analyze their voice.

[2058] Input: Condition data (JSON format) and user input

[2059] Output: Sentiment analysis result (e.g., user is stressed)

[2060] Step 4: Call the generative AI model

[2061] The server passes the analyzed emotional state and condition data to a generative AI model, which then generates an optimal product candidate list. The generative AI model performs analysis using big data.

[2062] Input: Condition data and emotion analysis results

[2063] Output: Product candidate list generated by the generative AI model

[2064] Step 5: Product candidate list formatting

[2065] The server formats the product candidate list obtained from the generative AI model and prepares it for return to the user.

[2066] Input: Product candidate list generated by the generative AI model

[2067] Output: A formatted list of product candidates

[2068] Step 6: Reply to the user

[2069] The server sends the formatted product candidate list to the user's chat screen.

[2070] Input: A formatted list of product candidates

[2071] Output: Product candidate list displayed on the user's chat screen

[2072] Step 7: User feedback

[2073] Users can review the provided product shortlist and enter additional preferences and feedback, such as "Mattress A looks good, but let me know if there's anything cheaper."

[2074] Input: User additional requirements and feedback

[2075] Output: User feedback sent from the terminal in text format

[2076] Step 8: Server reception and reparsing

[2077] The server receives feedback sent by the user and invokes a regenerative AI model that reflects condition modifications and emotional states.

[2078] Input: User feedback (text format)

[2079] Output: Reanalyzed condition data and sentiment analysis results

[2080] Step 9: Regenerate AI model call

[2081] The server then calls the generative AI model again based on the reanalyzed condition data and emotional state, and generates a new list of product candidates.

[2082] Input: Reanalyzed condition data and emotional state

[2083] Output: A new product candidate list generated by the generative AI model

[2084] Step 10: Provide a regenerated candidate list

[2085] The server formats the regenerated product candidate list and provides it to the user's chat screen again.

[2086] Input: A new product candidate list generated by the regenerative AI model

[2087] Output: New, re-formatted product candidate list

[2088] Step 11: Final user selection

[2089] The user finally selects the most suitable product from the presented product candidates and decides to "purchase Mattress C."

[2090] Input: User's final selection

[2091] Output: The final selected product information is sent to the server.

[2092] Step 12: Purchase procedure guide

[2093] The server receives the final selected product information and provides the user with detailed purchase procedure information and a link to the purchase page.

[2094] Input: User's last selected product information

[2095] Output: A purchase procedure message will be displayed on the user's chat screen.

[2096] By combining the above steps, users can efficiently select the most suitable mattress and proceed with the purchase process while taking into account their own emotional state.

[2097] (Application example 2)

[2098] 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."

[2099] Conventional product selection systems suggest products based on user condition data, but because they do not take the user's emotional state into account, they have the problem of not being able to make suggestions that are optimal for the user's psychological state. In addition, there is no way to display product location information within the store, making it difficult for users to search for products efficiently. This reduces users' motivation to purchase, and there are issues such as the time and effort required to select the most suitable product.

[2100] 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: an input means for inputting user condition data; a communication means for receiving the user condition data; a generation AI model means for analyzing the received condition data and emotion analysis data to generate an optimal product candidate list; a communication means for formatting the generated product candidate list and returning it to the user; a communication means for reflecting condition modifications based on user feedback and emotion analysis data and again calling the generation AI model means; an emotion recognition means for recognizing the user's emotional state in real time; and a visualization means for displaying product locations within a physical store for the user. This enables personalized product suggestions based on the user's emotional state, allowing the user to efficiently search for products within the store.

[2101] "User condition data" refers to the desired conditions (budget, hardness, size, brand, etc.) entered by the user.

[2102] "Input means" refers to a device or interface that allows a user to input condition data.

[2103] "Communication means" refers to a network interface for sending and receiving data.

[2104] "Emotion analysis data" is data that represents the user's emotional state after analyzing it.

[2105] "Generative AI model means" refers to an artificial intelligence model for generating an optimal product candidate list using received data.

[2106] "Emotion recognition means" refers to technology that analyzes a user's facial expressions and tone of voice to recognize their emotional state in real time.

[2107] "Visualization means" refers to displays and AR technologies that allow users to visually view information.

[2108] "Product Candidate List" means a list of optimal products generated by the generative AI model means and based on the user's criteria and emotional state.

[2109] "Feedback" refers to any additional preferences or comments provided by the User.

[2110] "Physical store" refers to an actual store space.

[2111] To implement this invention, a system including the following components is required. This system is designed to suggest optimal products based on a user's condition data and emotion data. Specifically, it uses a user input means, a communication means, an emotion recognition means, a generation AI model means, and a visualization means.

[2112] First, the user puts on the smart glasses and inputs their requirements by voice or touch. This includes information such as budget, hardness, size, and brand. For example, "budget is under ¥50,000, hardness is medium, size is single, and brand is not specified."

[2113] The input condition data is sent to the server via a communication means. The server receives the data and analyzes the user's emotional state via an emotion recognition means. Emotion recognition uses the smart glasses' camera and microphone to analyze the user's facial expressions and tone of voice. For example, it can determine whether the user is feeling stressed.

[2114] The server then passes the received condition data and emotion data to the generative AI model means, which analyzes this information using big data. The generative AI model generates an optimal product candidate list based on the condition and emotion state. The generated list is sent back to the smart glasses via the communication means.

[2115] The smart glasses use a visualization tool to present the generated product candidate list to the user, along with the product location information, allowing the user to efficiently search for products in the store.

[2116] For example, if "Mattress A" is included in the generated candidate list, its location information will be displayed on the display. The user can walk around the store while checking this information and check the products directly. If the user gives feedback such as "Mattress A looks good, but please let me know if there is anything cheaper," this feedback will also be sent to the server via the communication means.

[2117] The server then calls the generative AI model again based on the received feedback and emotional data, updating the product candidate list. In this way, the user's feedback and emotional state are reflected in real time, allowing for smooth selection of the most suitable product.

[2118] As a concrete example, the following prompt sentence is input to the generative AI model:

[2119] "List products (mattresses) that match the following criteria: Conditions: { "Budget": "50000", "Firmness": "Medium", "Size": "Single", "Brand": "Not Specified"} User emotional state: "Feeling stressed""

[2120] The system allows users to find the perfect mattress in an efficient and emotionally sensitive way.

[2121] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2122] Step 1:

[2123] The user puts on the smart glasses and inputs the desired conditions by voice or touch operation. The input conditions include budget, hardness, size, brand, etc. Data such as "budget within 50,000 yen, hardness medium, size single, no particular brand" is input. The input data is temporarily stored in the smart glasses' internal memory.

[2124] Step 2:

[2125] The user's input data is sent to the server via a communication method, either via the internet or the store's Wi-Fi. The server processes the received condition data as data for analysis. At this time, the received data is converted to JSON format and saved in the server's data storage.

[2126] Step 3:

[2127] The server then calls the emotion recognition unit based on the received data. It uses the smart glasses' camera and microphone to collect the user's facial expressions and tone of voice, analyzes them, and recognizes their emotional state. The analyzed emotional state is then classified into categories such as "feeling stressed" and output as data.

[2128] Step 4:

[2129] The server passes the condition data and the emotion data obtained by emotion recognition to the generative AI model means. The generative AI model receives these data as input and performs analysis using big data. The analysis process includes running database queries, filtering, and scoring. The generative AI model outputs a list of optimal product candidates, which is sent back to the server in JSON format.

[2130] Step 5:

[2131] The server then formats the received product candidate list and converts it into a data format for sending back to the user. This conversion process includes formatting data such as product name, price, and location information. The formatted data is then sent back to the smart glasses via a communication means.

[2132] Step 6:

[2133] The smart glasses display the received product candidate list to the user using a visualization means. The smart glasses display shows the names, prices, and location information of the candidate products in the store. The user confirms the displayed information and moves around the store to actually check the products.

[2134] Step 7:

[2135] After checking the products, the user can provide feedback by voice input or touch operation on the smart glasses. For example, they can input feedback such as, "Mattress A looks good, but please let me know if there is anything cheaper." This feedback is sent to the server via a communication means.

[2136] Step 8:

[2137] The server receives the feedback and updates the condition data and emotion data. Based on the updated data, it calls the generation AI model means again to generate a new product candidate list. This new candidate list is also shaped by the above process and sent to the smart glasses again.

[2138] Step 9:

[2139] The smart glasses will then display the new product candidate list using a visualization tool and present it to the user. The user will then review the products based on the information provided and make a final selection. After making their selection, they will be guided through the purchase process.

[2140] Through these steps, users can efficiently select the best mattress for them. This system provides personalized product recommendations based on the user's emotional state and efficient in-store navigation.

[2141] 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.

[2142] 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.

[2143] 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 robot 414.

[2144] 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.

[2145] 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.

[2146] 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.

[2147] 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).

[2148] 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.

[2149] 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."

[2150] 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.

[2151] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2152] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2153] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2154] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[2155] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[2156] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2157] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[2158] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[2159] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[2160] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[2161] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[2162] The following is further disclosed regarding the above embodiment.

[2163] (Claim 1)

[2164] A means for inputting user condition data;

[2165] a server means for receiving user condition data;

[2166] A generating AI model means for analyzing the big data based on the received condition data and generating an optimal product candidate list;

[2167] a server means for formatting the generated product candidate list and returning it to the user;

[2168] A server means for receiving feedback from a user, reflecting the condition modification, and calling the generation AI model means again;

[2169] A system including:

[2170] (Claim 2)

[2171] 2. The system according to claim 1, wherein a plurality of product candidates are presented based on condition data input by a user.

[2172] (Claim 3)

[2173] 10. The system of claim 1, wherein the product candidate list is regenerated based on feedback from the user.

[2174] "Example 1"

[2175] (Claim 1)

[2176] means for inputting user condition data;

[2177] data processing means for receiving user condition data;

[2178] a machine learning model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list;

[2179] a data processing means for processing the generated product candidate list and returning it to the user;

[2180] A data processing means for receiving feedback from users, reflecting the condition modifications, and calling the machine learning model means again;

[2181] A system including:

[2182] (Claim 2)

[2183] 2. The system according to claim 1, wherein a plurality of product candidates are presented based on condition data input by a user.

[2184] (Claim 3)

[2185] 10. The system of claim 1, wherein the product candidate list is regenerated based on feedback from the user.

[2186] "Application Example 1"

[2187] (Claim 1)

[2188] A means for inputting user condition data;

[2189] a server means for receiving user condition data;

[2190] A generation AI model means for analyzing large-scale data based on the received condition data and generating an optimal product candidate list;

[2191] a server means for formatting the generated product candidate list and returning it to the user;

[2192] A server means for receiving feedback from a user, reflecting the condition modification, and calling the generation AI model means again;

[2193] a user interface means using a smartphone;

[2194] A means for presenting the generated product candidate list to the user and providing feedback on additional conditions;

[2195] A system including:

[2196] (Claim 2)

[2197] 2. The system according to claim 1, wherein a plurality of product candidates are presented based on condition data input by a user.

[2198] (Claim 3)

[2199] 10. The system of claim 1, wherein the product candidate list is regenerated based on feedback from the user.

[2200] "Example 2: Combining Emotion Engines"

[2201] (Claim 1)

[2202] an input device means for inputting user condition data;

[2203] a server device means for receiving user condition data;

[2204] a generating AI model means for generating an optimal product candidate list based on the received condition data and the user's emotional state recognized by the emotion analysis means;

[2205] a server device means for formatting the generated product candidate list and returning it to the user;

[2206] A server device means for receiving feedback from a user, and for re-calling the generating AI model means by reflecting the condition modification and the emotional state;

[2207] A system including:

[2208] (Claim 2)

[2209] 10. The system of claim 1, wherein a plurality of product candidates are presented based on condition data entered by the user and the recognized emotional state.

[2210] (Claim 3)

[2211] 10. The system of claim 1, wherein the product candidate list is regenerated based on feedback from the user and the perceived emotional state.

[2212] "Application example 2 when combining emotion engines"

[2213] (Claim 1)

[2214] an input means for inputting user condition data;

[2215] a communication means for receiving user condition data;

[2216] A generating AI model means for analyzing the received condition data and sentiment analysis data and generating an optimal product candidate list;

[2217] A communication means for formatting the generated product candidate list and returning it to the user;

[2218] A communication means for calling the generation AI model means again by reflecting the condition modification based on the feedback from the user and the sentiment analysis data;

[2219] an emotion recognition means for recognizing the user's emotional state in real time;

[2220] a visualization means for users to view product locations within a physical store;

[2221] A system including:

[2222] (Claim 2)

[2223] 10. The system according to claim 1, wherein a plurality of product candidates are presented based on condition data and emotional state input by the user.

[2224] (Claim 3)

[2225] 10. The system of claim 1, wherein the product candidate list is regenerated based on feedback and emotional state from the user. [Explanation of symbols]

[2226] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for inputting user condition data; a server means for receiving user condition data; A generating AI model means for analyzing the big data based on the received condition data and generating an optimal product candidate list; a server means for formatting the generated product candidate list and returning it to the user; A server means for receiving feedback from a user, reflecting the condition modification, and calling the generation AI model means again; A system including:

2. 2. The system according to claim 1, wherein a plurality of product candidates are presented based on condition data input by a user.

3. The system of claim 1 , wherein the product candidate list is regenerated based on feedback from the user.

Images