System

Abstract

To provide a system.SOLUTION: A system includes: means for receiving a prompt sentence from a user; means for inputting the prompt sentence into a generative AI model to generate AI avatars that has learned characteristics of individual users; means for causing the AI avatars to talk; means for analyzing a log of the talk to calculate compatibility between the AI avatars; and means for notifying the user of a result of calculating the compatibility.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] Traditional matchmaking services have issues such as matchmakers' labor costs and time constraints, making them difficult for users to use. [Means for solving the problem]

[0005] To solve this problem, we provide a system that generates AI avatars that learn the characteristics of individual users and then matches the created AI avatars with each other to determine compatibility between users. Furthermore, we provide a means for users to visually check the progress of the matchmaking between the AI ​​avatars, and once the AI ​​avatars have been trained, they can be used any number of times, allowing for time-efficient matchmaking. [Brief explanation of the drawings]

[0006] [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. 2 is a sequence diagram showing a flow of processing in the data processing system according to the first embodiment of the first form example. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Embodiment 1. [Figure 13] FIG. 10 is a sequence diagram showing a processing flow of a data processing system in a second embodiment of the second form example. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Embodiment Example 2. [Figure 15] FIG. 10 is a sequence diagram showing the flow of processing in a data processing system according to a third embodiment of the third embodiment. [Figure 16]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Embodiment 3. [Figure 17] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when an emotion engine is combined. [Figure 19] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the second embodiment of the second form example when an emotion engine is combined. [Figure 20] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when an emotion engine is combined. [Figure 21] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the third embodiment of the third form example when an emotion engine is combined. [Figure 22] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0007] 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.

[0008] First, the terms used in the following description will be explained.

[0009] 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, the 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), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] 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.

[0011] 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.

[0012] 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.

[0013] 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."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 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.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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.

[0022] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0023] 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.

[0024] 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.

[0025] 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.

[0026] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0027] "Example 1"

[0028] In one embodiment of the present invention, a system is provided that includes an interface through which a user can input their own characteristics. This interface is designed to allow detailed input of individual characteristics, such as the user's hobbies, values, and lifestyle. The input characteristics are used as learning data for AI avatars, and each AI avatar reflects the user's characteristics.

[0029] "Example 2"

[0030] Next, the generated AI avatars are matched with each other. This matchmaking takes place in the form of a conversation between the AI ​​avatars, and compatibility is judged based on the content and progress of the conversation. Specifically, the flow of the conversation between the AI ​​avatars, choice of topics, speed of reaction, etc. are analyzed, and compatibility is calculated from these elements.

[0031] "Example 3"

[0032] Furthermore, users can visually check the progress of the matchmaking between the AI ​​avatars. Specifically, the conversation between the AI ​​avatars is displayed as text, and the avatars' facial expressions and movements are animated. This allows users to evaluate not only the compatibility between the AI ​​avatars, but also the process itself.

[0033] The processing flow of each embodiment will be described below.

[0034] "Example 1"

[0035] Step 1: The user accesses the system and enters their characteristics. Characteristics are items in which users can enter detailed information about their individual characteristics, such as hobbies, values, and lifestyle.

[0036] Step 2: The system generates an AI avatar based on the input characteristics. This AI avatar will reflect the user's characteristics.

[0037] "Example 2"

[0038] Step 1: The generated AI avatars begin meeting with each other. The meeting takes place in the form of a conversation between the AI ​​avatars.

[0039] Step 2: The system analyzes the flow of conversation between the AI ​​avatars, their choice of topics, and the speed of their reactions, and calculates compatibility based on these factors.

[0040] "Example 3"

[0041] Step 1: The user visually checks the matchmaking process between the AI ​​avatars through the system.

[0042] Step 2: Specifically, the content of conversations between AI avatars is displayed as text, and the avatars' facial expressions and movements are displayed as animations.

[0043] Step 3: This allows users to evaluate not only the compatibility between AI avatars, but also the process.

[0044] Example 1

[0045] Next, a description will be given of Example 1 of Form 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."

[0046] Conventional AI avatar generation systems have had difficulty efficiently generating AI avatars that reflect the characteristics of users and determining compatibility between users. Furthermore, there was a lack of a way to visually check the matchmaking process between the generated AI avatars, making it less convenient for users. Furthermore, there were also issues with the low reusability of AI avatars and poor time efficiency.

[0047] The identification process by the identification processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes a means having an interface for inputting the characteristics of individual users, a means for saving the input characteristics in a database, a means for creating a training dataset based on the saved data, a means for generating an AI avatar using a generative AI model, a means for matching the generated AI avatars with each other, and a means for determining the compatibility between users based on the results of the matchmaking. This makes it possible to efficiently generate AI avatars that reflect the characteristics of users and accurately determine the compatibility between users. In addition, the matchmaking process between the generated AI avatars can be visually confirmed, improving convenience for users. Furthermore, the reusability of AI avatars is increased, and time efficiency is improved.

[0048] "User" refers to an individual who uses the system to input their own characteristics and generate an AI avatar.

[0049] "Interface" refers to the means by which a screen or form is presented to a user to input their characteristics.

[0050] "Characteristics" refers to individual attributes and information such as a user's hobbies, values, and lifestyle.

[0051] "Database" refers to an information system for storing and managing input characteristics.

[0052] A "training dataset" refers to a collection of data used to train an AI model based on stored feature data.

[0053] A "generative AI model" refers to an algorithm or program for generating an AI avatar based on a training dataset.

[0054] An "AI avatar" refers to a virtual character generated to reflect the user's characteristics.

[0055] "Matchmaking" refers to the process by which generated AI avatars interact with each other and determine their compatibility.

[0056] "Compatibility" refers to an indicator that shows how closely users' characteristics and values ​​match.

[0057] "Means for visual confirmation" refers to a function that allows users to view the matchmaking process between the generated AI avatars on the screen.

[0058] "Reusability" refers to the property that an AI avatar that has been trained can be used any number of times.

[0059] "Time efficiency" refers to the speed at which the AI ​​avatar generation and matchmaking process is carried out.

[0060] The present invention provides a system having an interface for allowing a user to input his / her own characteristics. A specific embodiment of this system will be described below.

[0061] First, the user inputs their characteristics using an interface on a web browser. This interface is built using HTML, CSS, and JavaScript (registered trademark). The user can input detailed information about their individual characteristics, such as hobbies, values, and lifestyle. For example, the user might input "hobbies: reading, values: cherishing family, lifestyle: health-conscious."

[0062] The terminal then sends the data entered by the user to the server using an HTTP POST request. This communication uses Axios, a JavaScript HTTP client library. The terminal sends the entered data to the server in JSON format.

[0063] The server analyzes the received data and saves it in a database. A relational database such as MySQL (registered trademark) or PostgreSQL is used for the database. The server analyzes the received data and saves it in the database as "User ID: 123, Hobby: Reading, Values: Cherishing Family, Lifestyle: Health-Oriented."

[0064] Next, the server creates a training dataset based on the stored data. This dataset reflects the user's characteristics. The server creates a training dataset for the AI ​​model based on the data retrieved from the database.

[0065] The server uses the generated dataset to generate an AI avatar. This process uses a generative AI model (such as GPT-3 (registered trademark) or BERT). The server uses the generative AI model to generate an AI avatar with the following characteristics: Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious.

[0066] Finally, the device displays the AI ​​avatar received from the server. A front-end framework such as React.js is used for display. The device displays the AI ​​avatar received from the server using React.js, showing the user, "Hello, I'm a health-conscious avatar who loves reading and values ​​family."

[0067] As a specific example, if a user enters "Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious," the server will generate an AI avatar based on this data and display to the user, "Hello, I'm a health-conscious avatar who loves reading and cares for family."

[0068] An example prompt is, "Based on the user's hobbies, values, and lifestyle, please generate an AI avatar that reflects their individual characteristics. For example, if their hobby is reading, their values ​​are family-oriented, and their lifestyle is health-conscious, please create an AI avatar with those characteristics."

[0069] In this way, the system can efficiently generate AI avatars that reflect the characteristics of each user and determine compatibility between users. Furthermore, users can visually check the matchmaking process between the generated AI avatars, improving convenience for users. Furthermore, the reusability of AI avatars is enhanced, improving time efficiency.

[0070] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0071] Step 1:

[0072] User enters characteristics

[0073] A user inputs their own characteristics using an interface on a web browser. This interface is built using HTML, CSS, and JavaScript. Specifically, the user inputs "Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" into the form and clicks the submit button. The input data includes individual characteristics such as hobbies, values, and lifestyle.

[0074] Step 2:

[0075] The device sends the data to the server

[0076] The device sends the data entered by the user to the server using an HTTP POST request. This communication uses Axios, a JavaScript HTTP client library. Specifically, the device sends the data "Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" in JSON format to the server. The input is the feature data entered by the user, and the output is the JSON data sent to the server.

[0077] Step 3:

[0078] The server receives and stores the data

[0079] The server analyzes the received data and stores it in a database. A relational database such as MySQL or PostgreSQL is used for the database. Specifically, the server analyzes the received data and stores it in the database as "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-Oriented." The input is JSON data sent from the device, and the output is the user feature data stored in the database.

[0080] Step 4:

[0081] The server creates the training dataset

[0082] The server creates a training dataset based on the stored data. This dataset reflects the user's characteristics. Specifically, the server retrieves the following data from the database: "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" and creates a training dataset for the AI ​​model. The input is the user characteristic data retrieved from the database, and the output is the training dataset.

[0083] Step 5:

[0084] The server generates an AI avatar

[0085] The server uses the generated dataset to generate an AI avatar. This process uses a generative AI model (such as GPT-3 or BERT). Specifically, the server uses the generative AI model to generate an AI avatar with the following characteristics: Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious. The input is the training dataset, and the output is the generated AI avatar.

[0086] Step 6:

[0087] The device displays an AI avatar

[0088] The device displays the AI ​​avatar received from the server. A front-end framework such as React.js is used for display. Specifically, the device displays the AI ​​avatar received from the server using React.js, showing the user "Hello, I'm a health-conscious avatar who loves reading and values ​​family." The input is the AI ​​avatar received from the server, and the output is the AI ​​avatar displayed to the user.

[0089] (Application example 1)

[0090] Next, a description will be given of Application Example 1 of Embodiment 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."

[0091] Conventional content distribution services have the problem of not providing sufficient personalized content recommendations based on users' hobbies, values, and lifestyles. Furthermore, there is a lack of systems that allow users to input their own detailed characteristics and then recommend the most appropriate content based on that information, which can lead to a decline in user satisfaction.

[0092] 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.

[0093] In this invention, the server includes a means for generating an AI avatar that has learned the characteristics of each user, a means for recommending optimal content to the user based on the generated AI avatar, and a means for providing the recommended content to the user. This enables personalized content recommendations based on the user's characteristics, thereby improving user satisfaction.

[0094] "User" refers to any individual who uses the system.

[0095] "Characteristics" refers to individual attributes such as a user's hobbies, values, and lifestyle.

[0096] An "AI avatar" refers to an artificial intelligence model that learns the user's characteristics and is generated based on them.

[0097] "Content" refers to information and entertainment such as movies, music, and articles.

[0098] "Recommendation" refers to the act of selecting and presenting the most suitable content based on the user's characteristics.

[0099] "Providing" refers to the act of actually showing or making available the recommended content to a user.

[0100] "Visual confirmation means" refers to a method that allows users to visually confirm the recommendation process and results through a screen or other means.

[0101] "Time-efficient" refers to a state in which a goal can be achieved effectively in a short amount of time.

[0102] The system for implementing this invention comprises a server and a user terminal. The server includes a means for generating an AI avatar that has learned the characteristics of each user, a means for recommending optimal content to the user based on the generated AI avatar, and a means for providing the recommended content to the user.

[0103] Program processing explanation

[0104] 1. User characteristics input

[0105] The user terminal provides an interface for users to input their own characteristics (hobbies, values, lifestyle, etc.). This interface is implemented as a smartphone application. The characteristics input by the user are sent to the server.

[0106] 2. AI avatar generation

[0107] The server generates an AI avatar based on the received user feature data. This AI avatar is trained using a generative AI model. Specifically, it uses Python and the scikit-learn library to vectorize the user's features and save them as an AI avatar.

[0108] 3. Content Recommendations

[0109] The server then recommends the most suitable content to the user based on the generated AI avatar. The recommendation algorithm is implemented using TF-IDF vectorization and cosine similarity calculation, which selects the content that most closely matches the user's characteristics.

[0110] 4. Provision of Content

[0111] The server transmits the recommended content to the user terminal for visual confirmation by the user, and the user terminal provides an interface for displaying the recommended content.

[0112] Specific examples

[0113] For example, if a user enters the following characteristics:

[0114] Hobbies: Watching movies, listening to music

[0115] Values: Cherishing time with family

[0116] Lifestyle: Health-conscious

[0117] In this case, an example prompt to input to the generative AI model would be:

[0118] The user's hobbies are watching movies and listening to music, and their values ​​include spending time with family. Their lifestyle is health-conscious. Please recommend the most suitable content for this user.

[0119] Based on this prompt, the server recommends the most suitable content for the user, such as movies, music, or articles, and provides it to the user's device, allowing the user to enjoy personalized content based on their own characteristics.

[0120] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0121] Step 1:

[0122] The user device provides an interface where the user can input their own characteristics (hobbies, values, lifestyle, etc.). The characteristics input by the user are sent to the server in JSON format. Examples of input data include "Hobbies: watching movies, listening to music," "Values: valuing time with family," and "Lifestyle: health-conscious."

[0123] Step 2:

[0124] The server analyzes the received user feature data and generates an AI avatar using a generative AI model. Specifically, it uses Python and the scikit-learn library to vectorize the user features using TF-IDF and save them as an AI avatar. The input data are the user features, and the output data is the vectorized AI avatar.

[0125] Step 3:

[0126] The server recommends the most suitable content to the user based on the generated AI avatar. The recommendation algorithm is implemented using TF-IDF vectorization and cosine similarity calculation. Specifically, it compares the user's feature vector with the vector in the content list and selects the content with the highest similarity. The input data is the AI ​​avatar and the content list, and the output data is the recommended content.

[0127] Step 4:

[0128] The server transmits the recommended content to the user terminal. The user terminal provides an interface for displaying the recommended content. Specifically, the recommended content, such as movies, music, or articles, is displayed on the user's screen. The input data is the recommended content, and the output data is the content displayed on the user terminal.

[0129] Step 5:

[0130] The user can visually check and use the recommended content. When the user selects content, the selection information is fed back to the server and used for the next recommendation. The input data is the user's selection information, and the output data is the feedback information that will be reflected in the next recommendation.

[0131] Example 2

[0132] Next, a description will be given of Example 2 of Form 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."

[0133] Conventional matchmaking systems using AI avatars have had difficulty accurately determining compatibility between users. Furthermore, they lacked the functionality to analyze the content and progress of conversations in real time and calculate compatibility, making it difficult for users to obtain satisfactory results. Furthermore, there was no way to visually check the matchmaking process between the generated AI avatars, making it difficult for users to understand or accept the system.

[0134] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0135] In this invention, the server includes a means for a user to input a prompt sentence, a means for generating an AI avatar using a generative AI model, a means for matching the generated AI avatars, a means for analyzing the content and progress of the conversation, a means for calculating the compatibility between the AI ​​avatars, a means for notifying the user of the compatibility results, a means for the user to visually confirm the progress of the matchmaking between the AI ​​avatars, and a means for allowing the AI ​​avatar to be used repeatedly once trained, thereby allowing the matchmaking to proceed time-efficiently. This makes it possible to accurately determine the compatibility between users and provide satisfactory results. Furthermore, since users can visually confirm the progress of the matchmaking, understanding and satisfaction are improved.

[0136] A "prompt sentence" is an instruction sentence that a user inputs to a generative AI model to generate an AI avatar.

[0137] A "generative AI model" is an artificial intelligence model for generating an AI avatar based on an input prompt sentence.

[0138] An "AI avatar" is a virtual character generated by a generative AI model with specific characteristics and personality.

[0139] "Matchmaking" is a process in which generated AI avatars converse with each other to determine their compatibility.

[0140] "Means for analyzing the content and progress of conversations" refers to a function that monitors and evaluates in real time the flow of conversations between AI avatars, topic selection, speed of reactions, etc.

[0141] The "means for calculating compatibility" is a function that calculates the compatibility between AI avatars in the form of numbers, graphs, etc. based on the results of conversation analysis.

[0142] The "means for notifying compatibility results" is a function for notifying the user of the calculated compatibility results.

[0143] "Visual confirmation means" is a function that allows users to visually check the progress of matchmaking between AI avatars.

[0144] "A time-efficient means of conducting matchmaking" is a function that allows the AI ​​avatar to progress through matchmaking in a short amount of time, and can be used any number of times once it has been trained.

[0145] This invention is a system in which a user inputs a prompt sentence, generates an AI avatar using a generative AI model, and matches the generated AI avatars with each other. Specific embodiments of this system are described below.

[0146] First, the user inputs a prompt using a terminal. The prompt is an instruction for the generative AI model to generate an AI avatar, such as "You are a 30-year-old man whose hobbies are reading and hiking. Please introduce yourself."

[0147] The server then receives the prompt and generates an AI avatar using a generative AI model (e.g., OpenAI's GPT-4 or Google's BERT). The generated AI avatar has information such as a self-introduction and hobbies based on the prompt.

[0148] The server pairs the generated AI avatars and initiates matchmaking. Matchmaking takes place by having the AI ​​avatars converse with each other. The server monitors the content and progress of the conversation in real time and analyzes the following factors:

[0149] Conversation flow: Check that the conversation is flowing smoothly, i.e., that the topic is kept on topic and that the conversation is unfolding naturally.

[0150] Topic selection: Evaluate whether the topics chosen by the AI ​​avatar are appropriate, for example, ensuring that topics are based on shared interests.

[0151] Reaction speed: Measures how quickly the AI ​​avatar responds to what the other person says. Evaluates whether it responds quickly and appropriately.

[0152] The server analyzes these factors comprehensively and calculates the compatibility between the AI ​​avatars. The compatibility is displayed in the form of numbers and graphs. For example, if the compatibility is evaluated as 80%, the server notifies the user of the result.

[0153] Furthermore, the server provides a means for users to visually check the matchmaking process between the AI ​​avatars, allowing them to intuitively understand the progress and results of the matchmaking.

[0154] Furthermore, once the AI ​​avatar has been trained, it can be used any number of times, allowing for time-efficient matchmaking, enabling users to conduct multiple matchmaking sessions in a short amount of time and find the perfect partner.

[0155] The above is a specific embodiment for carrying out the present invention.

[0156] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0157] Step 1:

[0158] The user enters a prompt statement.

[0159] The user uses a terminal to input a prompt sentence to the generative AI model to generate an AI avatar. For example, the sentence might be, "You are a 30-year-old man whose hobbies are reading and hiking. Please introduce yourself." The input prompt sentence is sent to the server.

[0160] Step 2:

[0161] The server generates an AI avatar using the generative AI model.

[0162] The server receives the prompt sent by the user and inputs it into a generative AI model (e.g., OpenAI's GPT-4). The generative AI model generates an AI avatar based on the prompt and determines the avatar's characteristics and personality. The data of the generated AI avatar is stored on the server.

[0163] Step 3:

[0164] The server matches the generated AI avatars with each other.

[0165] The server pairs the generated AI avatars and initiates matchmaking. Matchmaking takes place in the form of conversation between the AI ​​avatars. The server instructs each AI avatar to start a conversation and collects conversation logs in real time.

[0166] Step 4:

[0167] The server analyzes the content and progress of the conversation.

[0168] The server analyzes the collected conversation logs and evaluates the following factors:

[0169] Conversation flow: Check that the conversation is flowing smoothly, i.e., that the topic is kept on topic and that the conversation is unfolding naturally.

[0170] Topic selection: Evaluate whether the topics chosen by the AI ​​avatar are appropriate, for example, ensuring that topics are based on shared interests.

[0171] Reaction speed: Measures how quickly the AI ​​avatar responds to what the other person says. Evaluates whether it responds quickly and appropriately.

[0172] Step 5:

[0173] The server calculates the compatibility between AI avatars.

[0174] The server calculates the compatibility between the AI ​​avatars based on the data collected in step 4. The compatibility is displayed in the form of a number or a graph. For example, if the compatibility is evaluated as 80%, the server generates the result.

[0175] Step 6:

[0176] The server notifies the user of the compatibility results.

[0177] The server notifies the user of the calculated compatibility results, which the user can check via their device. This allows the user to decide their next action based on the results of the matchmaking.

[0178] Step 7:

[0179] To provide a means for a server to visually check the state of a matchmaking session.

[0180] The server provides an interface that allows users to visually check the matchmaking process between AI avatars, and users can intuitively understand the progress and results of the matchmaking process through their devices.

[0181] Step 8:

[0182] The server makes the AI ​​avatars reusable.

[0183] The server allows the generated AI avatar to be trained once and then used again and again, allowing users to conduct multiple matchmaking sessions in a short amount of time and find the perfect partner.

[0184] (Application example 2)

[0185] Next, a description will be given of Application Example 2 of Form 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."

[0186] Conventional matchmaking systems using AI avatars have had difficulty accurately determining compatibility between users. They also lack the ability to take into account detailed factors such as the flow of conversation, topic selection, and speed of response, resulting in insufficient calculations of compatibility. Furthermore, there was a lack of a way for users to visually check the progress of the matchmaking, which hindered the user experience.

[0187] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining the compatibility between users based on the results of the matchmaking, means for analyzing the flow of conversation, topic selection, and reaction speed, and means for calculating compatibility. This makes it possible to more accurately determine the compatibility between users and visually check the matchmaking process.

[0188] "Means for generating an AI avatar that has learned the characteristics of each individual user" refers to a method or device for learning the individual characteristics and preferences of each user and generating an AI avatar based on that.

[0189] "Means for matching generated AI avatars" refers to a method or device for matching generated AI avatars through conversation.

[0190] "Means for determining compatibility between users based on matchmaking results" refers to a method or device for analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users based on the results.

[0191] "Means for analyzing conversation flow, topic selection, and reaction speed" refers to methods and devices for analyzing the flow of conversations between AI avatars, the topics selected, reaction speed, etc.

[0192] The "means for calculating compatibility" refers to a method or device for quantifying the compatibility between users based on the results of conversation analysis.

[0193] "Means for visual confirmation" refers to a method or device that allows users to visually confirm the matchmaking process between AI avatars.

[0194] "A means for efficiently conducting matchmaking in a time-efficient manner" refers to a method or device that allows an AI avatar to be trained once and then used any number of times, allowing matchmaking to proceed efficiently in a short amount of time.

[0195] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The following explains how each element functions.

[0196] First, the server has a means for generating an AI avatar that has learned the characteristics of each individual user. This means generates an AI avatar using a generative AI model based on the user's input data (e.g., profile information and past conversation history). The generated AI avatar reflects the user's characteristics and has individual preferences and personality.

[0197] Next, the server has a means for matching the generated AI avatars with each other. This means generates prompts for the generated AI avatars to converse with each other, and simulates the conversation using the generative AI model. For example, the following prompts are used:

[0198] Avatar 1: Hello, how are you doing these days?

[0199] Avatar 2: I'm fine. How about you?

[0200] Start a conversation.

[0201] The server has a means for analyzing the flow of conversation, topic selection, and speed of response. This means analyzes the generated conversation text and evaluates elements such as the flow of conversation, topic selection, and speed of response. This allows data to be obtained for quantifying the quality of the conversation and compatibility.

[0202] Furthermore, the server has a means for calculating compatibility, which quantifies the compatibility between users based on the results of conversation analysis. For example, if the conversation flows smoothly and the choice of topics reflects common interests, a high compatibility score is calculated.

[0203] The device has a means for users to visually check the matchmaking process. This allows users to visually check the conversation between the AI ​​avatars in real time. For example, they can use a smartphone or head-mounted display to visually experience the conversation in a virtual space.

[0204] Users can control the AI ​​avatar through their device and check the results of their matchmaking. Users can customize their own avatar and check compatibility by talking with other users' avatars.

[0205] This system allows users to more accurately judge compatibility between each other and visually check the matchmaking process. Specifically, it provides a realistic conversation experience by simulating conversations using a generative AI model and generating conversations using prompts.

[0206] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0207] Step 1:

[0208] The server receives input data from the user. The input data includes the user's profile information and past conversation history. Based on this data, a generative AI model is used to generate an AI avatar that learns the user's characteristics. The generated AI avatar reflects the user's individual preferences and personality. The input is the user's data, and the output is the generated AI avatar.

[0209] Step 2:

[0210] The server generates a prompt to match the generated AI avatars. The prompt is text to start a conversation between the AI ​​avatars. For example, it takes the form "Avatar 1: Hello, how are you?\nAvatar 2: I'm fine. How about you?\nLet's start the conversation." The input is the generated AI avatar, and the output is the prompt.

[0211] Step 3:

[0212] The server uses a generative AI model to simulate a conversation based on the prompt sentence. The generative AI model receives the prompt sentence as input and generates a conversation between AI avatars. The generated conversation is output in text format. The input is the prompt sentence, and the output is the generated conversation text.

[0213] Step 4:

[0214] The server analyzes the generated conversation text. The analysis includes elements such as the flow of the conversation, topic selection, and speed of response. These elements are evaluated to obtain data to quantify the quality of the conversation and compatibility. The input is the generated conversation text, and the output is the analysis results.

[0215] Step 5:

[0216] The server calculates the compatibility between users based on the analysis results. Compatibility is high when the conversation flows smoothly and the topic choices reflect common interests. Compatibility is quantified and fed back to the user. The input is the analysis results, and the output is a compatibility score.

[0217] Step 6:

[0218] The terminal allows the user to visually check the progress of the matchmaking. The terminal visually displays the generated conversation text, allowing the user to check the conversation in real time. The input is the generated conversation text, and the output is the visual display.

[0219] Step 7:

[0220] Users control the AI ​​avatar through their device and check the matchmaking results. Users can customize their own avatar and check compatibility by conversing with other users' avatars. The input is a compatibility score, and the output is user feedback.

[0221] Example 3

[0222] Next, a description will be given of a third embodiment of the third embodiment. 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."

[0223] In conventional systems using AI avatars, there were limited means for determining compatibility between users, and the system was unable to visually confirm it. It was also difficult to evaluate the process of dialogue between AI avatars, making it less convenient for users. Furthermore, generating AI avatars and simulating dialogue took time, making it difficult to proceed efficiently.

[0224] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0225] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for having the generated AI avatars converse with each other, means for determining compatibility between users based on the results of the conversation, means for the user to input a prompt sentence, means for simulating a conversation between the generated AI avatars, means for generating animation data of the conversation content and the avatars, means for transmitting the generated data to a terminal, means for the terminal to display the conversation content and avatar animation, and means for the user to evaluate the compatibility between the AI ​​avatars and the progress of the conversation. This allows the user to visually check the state of the conversation between the AI ​​avatars and evaluate the process. Furthermore, once the AI ​​avatars have been trained, they can be used repeatedly, allowing for time-efficient conversations.

[0226] An "artificial intelligence avatar" is a virtual character that learns the user's characteristics and is generated to simulate conversations and actions.

[0227] A "prompt" is a text of instructions or questions entered by a user, which serves as the basis for generating an AI avatar and simulating a dialogue.

[0228] "Dialogue" refers to conversations and communication between artificial intelligence avatars, and is simulated based on user input.

[0229] "Simulation" is the process of virtually recreating the interactions and behaviors of artificially intelligent avatars, and is done using a generative AI model.

[0230] "Animation data" is data that visually expresses the facial expressions and movements of an artificial intelligence avatar, and is displayed to the user in real time.

[0231] A "terminal" is a device that allows a user to input prompts and visually check the dialogue and animation of an artificial intelligence avatar.

[0232] A "generative AI model" is a machine learning model used to generate artificial intelligence avatars and simulate dialogue, and includes, for example, natural language processing models.

[0233] "Compatibility" refers to the suitability and compatibility between users, as determined based on the results of conversations between AI avatars.

[0234] "Evaluation" refers to the judgment or feedback that the user gives based on the process and results of the dialogue between AI avatars.

[0235] The present invention is a system that uses an artificial intelligence avatar and provides a means for determining compatibility between users. A specific embodiment of this system will be described below.

[0236] First, the user inputs a prompt using the terminal. The prompt specifies the theme and content of the conversation between the AI ​​avatars. For example, the user can input a prompt such as "Start a matchmaking session between AI avatars discussing hobbies."

[0237] The server then receives the prompt and generates an AI avatar using a generative AI model, such as the natural language processing model GPT-4. The generated AI avatar is configured to have different personalities and hobbies based on the prompt.

[0238] The server simulates a conversation between the generated AI avatars. The content of the simulated conversation is generated in text format. For example, Avatar A might say, "I like reading. I especially like mystery novels," and Avatar B might reply, "I like reading, too. I've been hooked on historical novels lately."

[0239] Furthermore, the server generates animation data for the avatars' facial expressions and movements based on the content of the conversation. For example, it generates an animation of avatar A smiling and talking, and avatar B nodding.

[0240] The generated text data and animation data are sent from the server to the device. Based on the received data, the device displays the dialogue content as text to the user and animates the avatars' facial expressions and movements. This allows the user to visually confirm the dialogue between the AI ​​avatars.

[0241] Finally, the user can evaluate the compatibility and dialogue process between the AI ​​avatars based on the dialogue content and animation displayed on the device. For example, they can evaluate that "Avatar A and Avatar B seem to get along well."

[0242] This system allows users to visually check the state of conversations between AI avatars and evaluate the process. Furthermore, once the AI ​​avatars have been trained, they can be used any number of times, allowing for time-efficient conversations.

[0243] As an example, the following prompt sentence is used:

[0244] "Starting matchmaking with AI avatars that discuss hobbies"

[0245] "Starting a matchmaking session with AI avatars to discuss travel"

[0246] "We're starting a matchmaking session for AI avatars that discuss music."

[0247] The above is an embodiment of the present invention. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0248] Step 1:

[0249] The user enters a prompt statement.

[0250] The user enters a prompt into the input field of the device, for example, "Start a matchmaking session for an AI avatar to discuss hobbies." The entered prompt is sent from the device to the server.

[0251] Step 2:

[0252] The server receives the prompt sentence and generates an artificial intelligence avatar using a generative AI model.

[0253] The server receives a prompt sent by the user. It then uses a generative AI model (e.g., a natural language processing model) to generate two AI avatars based on the prompt. The generated avatars are configured to have different personalities and hobbies. The input is the prompt, and the output is the generated AI avatars.

[0254] Step 3:

[0255] The server simulates conversations between artificial intelligence avatars.

[0256] The server simulates a conversation between the generated AI avatars. Using the generative AI model, Avatar A and Avatar B have a conversation based on a prompt. For example, Avatar A says, "I like reading," and Avatar B replies, "I like reading, too." The input is the generated AI avatar, and the output is the simulated conversation content.

[0257] Step 4:

[0258] The server generates the conversation content and avatar animation data.

[0259] The server generates the simulated conversation content in text format. It also generates the avatars' facial expressions and movements as animation data. For example, it generates an animation of avatar A smiling and talking, and avatar B nodding. The input is the simulated conversation content, and the output is text data and animation data.

[0260] Step 5:

[0261] The server transmits the generated data to the terminal.

[0262] The server sends the generated text data and animation data of the conversation to the terminal. The input is the text data and animation data, and the output is sent to the terminal.

[0263] Step 6:

[0264] The device displays the conversation and an animated avatar.

[0265] The device displays the conversation content as text to the user based on the data received from the server. It also displays the avatar's facial expressions and movements as animation. For example, avatar A and avatar B are displayed on the screen, and the conversation content and facial expressions of each are displayed in real time. The input is the data received from the server, and the output is what is displayed to the user.

[0266] Step 7:

[0267] Users evaluate the compatibility between AI avatars and the conversation process.

[0268] The user checks the conversation content and animation displayed on the device and evaluates the compatibility between the AI ​​avatars and the progress of the conversation. For example, the user may evaluate that "Avatar A and Avatar B seem to get along well." The input is the displayed conversation content and animation, and the output is the user's evaluation.

[0269] (Application example 3)

[0270] Next, a description will be given of Application Example 3 of Form Example 3. 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."

[0271] Conventional AI avatar systems have limited means for determining compatibility between users, and users cannot visually check the interactions between AI avatars. Furthermore, when introducing or comparing products, they lack the interactive functionality to allow users to ask questions in real time and receive answers. This has led to problems such as reduced user satisfaction and convenience.

[0272] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0273] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for the AI ​​avatars to introduce and compare products with each other, means for visually checking this situation, and means for the user to ask the AI ​​avatar questions and receive responses in real time, thereby enabling users to interactively obtain product information while visually checking the interactions between the AI ​​avatars.

[0274] An "AI avatar" is a virtual character that learns the characteristics of each individual user and is generated based on those characteristics.

[0275] "Matchmaking" is a process in which generated AI avatars converse and interact with each other, and their compatibility is determined based on the results.

[0276] The "means of determining compatibility" is a method for analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users.

[0277] "Means for introducing and comparing products" is a function that allows an AI avatar to provide information about specific products and compare multiple products.

[0278] "Visual confirmation" refers to a method that allows users to visually observe the interactions and animations between AI avatars.

[0279] "Real-time reception" is a feature that allows users to ask questions to an AI avatar and receive answers instantly.

[0280] A system for implementing the present invention includes a server, a user terminal, and an AI avatar generation program. Specific embodiments are described below.

[0281] System Configuration

[0282] 1. Server:

[0283] Hardware: A server equipped with a high-performance CPU, GPU, and memory.

[0284] Software: TENSORFLOW (registered trademark) (generative AI model), Unity (animation display), OpenCV (image processing).

[0285] 2. User Device:

[0286] Hardware: Smartphone, Head-Mounted Display (HMD).

[0287] Software: Dedicated application (Unity-based).

[0288] Program processing

[0289] 1. AI Avatar Generation:

[0290] The server uses TensorFlow to build a generative AI model and have it learn the user's characteristics.

[0291] When a user enters a prompt sentence, the generative AI model generates an AI avatar based on its features.

[0292] 2. Conducting matchmaking:

[0293] The server matches the generated AI avatars with each other.

[0294] Based on the results of the matchmaking, the server determines the compatibility between users.

[0295] 3. Product introduction and comparison:

[0296] The server generates scenarios in which AI avatars introduce and compare products.

[0297] The user's device uses Unity to display the AI ​​avatar's facial expressions and movements in animated form.

[0298] 4. Visual Verification:

[0299] The user device uses OpenCV to track the user's gaze and facial expressions in real time, providing an interactive experience.

[0300] Users can visually see the interactions between AI avatars.

[0301] 5. Real-time Questions and Answers:

[0302] Users can ask questions to the AI ​​avatar.

[0303] The server generates an answer to the question in real time and transmits it to the user terminal.

[0304] Specific examples

[0305] The user puts on a smartphone or HMD and accesses a virtual store. An AI avatar appears on the screen, and the user enters a prompt: "Tell me about the features of my new smartphone." The AI ​​avatar responds to the question by comparing the features of several smartphones. The conversation is displayed as text, and the avatar's facial expressions and movements are animated, allowing the user to visually confirm the conversation.

[0306] Prompt Sentence Examples

[0307] "What's the difference between the latest smartphones and tablets?"

[0308] "Write a review for this product"

[0309] How does this product compare to other products?

[0310] In this way, users can interactively obtain product information while visually checking the interactions between AI avatars.

[0311] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0312] Step 1:

[0313] The server receives the user's feature data and uses TensorFlow to build a generative AI model. The input is the user's feature data, and the output is an AI avatar corresponding to the user. Specifically, the server analyzes the user's profile information and past behavioral data and generates the AI ​​avatar based on that.

[0314] Step 2:

[0315] The server matches the generated AI avatars with each other. The input is the generated multiple AI avatars, and the output is the matchmaking results data. Specifically, the AI ​​avatars converse with each other, and the content of the conversation is analyzed to evaluate compatibility.

[0316] Step 3:

[0317] The server determines the compatibility between users based on the matchmaking results. The input is matchmaking result data, and the output is compatibility evaluation data. Specifically, it analyzes emotions in the conversation and extracts commonalities to calculate a compatibility score.

[0318] Step 4:

[0319] The server generates scenarios in which AI avatars introduce and compare products. The input is the user's prompt, and the output is a product introduction or comparison scenario. Specifically, when a user inputs "Tell me about the features of my new smartphone," the AI ​​avatar generates a scenario that explains the product's features based on the question.

[0320] Step 5:

[0321] The device uses Unity to display animated expressions and movements of an AI avatar. The input is a product introduction or comparison scenario, and the output is an animated display. Specifically, the device renders the expressions and movements of the AI ​​avatar in real time as it explains the product's features.

[0322] Step 6:

[0323] The device uses OpenCV to track the user's gaze and facial expressions in real time, providing an interactive experience. The input is the user's gaze and facial expression data, and the output is interactive feedback. Specifically, an AI avatar provides detailed explanations of products that the user shows interest in.

[0324] Step 7:

[0325] Users can ask questions to the AI ​​avatar. The input is the user's question, and the output is the AI ​​avatar's answer. Specifically, when a user asks, "Please tell me a review of this product," the server generates an answer to the question and sends it to the device.

[0326] Step 8:

[0327] The server generates an answer to the question in real time and sends it to the user's device. The input is the user's question, and the output is the answer data. Specifically, the server uses a generative AI model to generate the optimal answer to the question and provides it to the user.

[0328] 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.

[0329] "Example 1"

[0330] One embodiment of the present invention is a means for generating an AI avatar incorporating an emotion engine. This emotion engine recognizes the user's emotions and reflects them in the AI ​​avatar's reactions. Specifically, when the user expresses emotions such as joy or sadness to the system, the emotion engine recognizes this and the AI ​​avatar responds accordingly. For example, if the user expresses joy, the AI ​​avatar will respond by smiling.

[0331] "Example 2"

[0332] Also, a method for matching AI avatars using an emotion engine is one embodiment of the present invention. When AI avatars are matched, the emotion engine recognizes the emotions shown by each AI avatar and determines the outcome of the match based on that. For example, if one AI avatar shows joy and the other AI avatar also shows joy, it is determined that they are compatible.

[0333] "Example 3"

[0334] Furthermore, one embodiment of the present invention also provides a means for providing feedback to the user on the AI ​​avatar's reactions using an emotion engine. The emotional reactions of the AI ​​avatar are visually fed back to the user. Specifically, if the AI ​​avatar shows joy, that reaction is displayed as a smile on the screen and fed back to the user. This allows the user to directly see how their own emotions are reflected in the AI ​​avatar.

[0335] The processing flow of each embodiment will be described below.

[0336] "Example 1"

[0337] Step 1: The user expresses their feelings to the system, for example, expressing emotions such as happiness or sadness through text, voice, or facial expressions.

[0338] Step 2: The emotion engine recognizes the emotion expressed by the user. This emotion recognition is done using technologies such as text analysis, speech analysis, and facial expression analysis.

[0339] Step 3: The AI ​​avatar receives input from the emotion engine and generates a corresponding reaction. For example, if the user expresses joy, the AI ​​avatar will respond by smiling.

[0340] "Example 2"

[0341] Step 1: The AI ​​avatars begin meeting with each other. During this meeting, each AI avatar will engage in conversation that reflects the user's characteristics.

[0342] Step 2: The emotion engine recognizes the emotions expressed by each AI avatar by analyzing the text, voice, and facial expressions generated by the AI ​​avatar.

[0343] Step 3: The emotion engine determines the outcome of the matchmaking based on the emotions recognized. For example, if one AI avatar shows joy and the other AI avatar also shows joy, it is determined that they are a good match.

[0344] "Example 3"

[0345] Step 1: The AI ​​avatar generates an emotional response. This emotional response is generated based on the user's emotions and the results of the matchmaking between the AI ​​avatars.

[0346] Step 2: The AI ​​avatar's emotional response is visually displayed. Specifically, if the AI ​​avatar expresses joy, that response is displayed as a smile on the screen.

[0347] Step 3: The displayed emotional response of the AI ​​avatar is fed back to the user, allowing the user to directly see how their own emotions are reflected in the AI ​​avatar.

[0348] Example 1

[0349] Next, a description will be given of Example 1 of Form 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."

[0350] Conventional AI avatar systems have difficulty fully reflecting the characteristics and emotions of users, and lack the means to determine the compatibility between generated AI avatars. This makes it impossible to accurately determine the compatibility between users, reducing the usefulness of the system.

[0351] 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.

[0352] In this invention, the server includes means for providing an interface for users to input their own characteristics, means for preprocessing the input characteristic data, means for generating an AI avatar using the preprocessed data, means for providing an interface for inputting user emotion data, means for analyzing the input emotion data and reflecting it in the AI ​​avatar, means for matching the generated AI avatars with each other, and means for determining compatibility between users based on the results of the matchmaking. This makes it possible to generate AI avatars that reflect the users' characteristics and emotions, and to accurately determine compatibility between users through matching between AI avatars.

[0353] "User" refers to an individual who uses the system to input their own characteristics and emotions to generate an AI avatar.

[0354] "Interface" refers to the part of a system that provides a screen or input means for users to input their characteristics and emotions.

[0355] "Characteristic data" refers to individual pieces of information entered by users, such as hobbies, values, and lifestyle.

[0356] "Preprocessing" refers to the process of converting input feature data into a format suitable for learning by an AI avatar.

[0357] An "AI avatar" is a virtual character that is generated to reflect the user's characteristics and emotions.

[0358] "Emotional data" refers to information about emotions such as joy, sadness, and anger entered by the user.

[0359] An "emotion engine" refers to software that analyzes input emotional data and reflects it in the reactions of an AI avatar.

[0360] "Matchmaking" refers to the process of having generated AI avatars interact with each other to determine their compatibility.

[0361] "Compatibility assessment" refers to the process of evaluating the compatibility between users based on the results of matchmaking.

[0362] This invention relates to a system that allows users to input their own characteristics and emotions, generates AI avatars based on the input, and matches the AI ​​avatars with each other. Specific embodiments of this system are described below.

[0363] System configuration

[0364] This system consists of three main elements: a server, a terminal, and a user.

[0365] server

[0366] The server has the following functions:

[0367] 1. Interface provision function: Provides an interface for users to input their own characteristics.

[0368] 2. Data preprocessing function: Preprocesses the input feature data and converts it into a format suitable for learning by the AI ​​avatar.

[0369] 3. AI avatar generation function: Generates an AI avatar using preprocessed data.

[0370] 4. Emotional data analysis function: Analyzes the user's emotional data and reflects it in the AI ​​avatar.

[0371] 5. Matchmaking function: Matchmaking between generated AI avatars.

[0372] 6. Compatibility assessment function: Determines the compatibility between users based on the results of matchmaking.

[0373] Terminal

[0374] A terminal is a device through which a user inputs feature data and emotion data. Specifically, this includes a PC, smartphone, tablet, etc.

[0375] User

[0376] Users are individuals who use the system to input their own characteristics and emotions to generate an AI avatar.

[0377] Program processing

[0378] The server receives and preprocesses the feature data entered by the user through the device. The preprocessed data is then used to generate an AI avatar using a generative AI model. The generated AI avatar reflects the user's features and, by incorporating an emotion engine, reacts according to the user's emotional data.

[0379] Specifically, the server does the following:

[0380] 1. Data collection: The characteristic data entered by the user through the terminal is sent to the server.

[0381] 2. Data processing: The server preprocesses the collected feature data and converts it into a format suitable for learning the AI ​​avatar.

[0382] 3. AI avatar generation: The server generates an AI avatar using the preprocessed data, incorporating an emotion engine.

[0383] 4. Emotion recognition and response: The server uses an emotion engine to recognize the user's emotions and reflect them in the AI ​​avatar.

[0384] Specific examples

[0385] Here are some examples of specific prompts:

[0386] Example prompt sentence:

[0387] Enter your user characteristics:

[0388] Hobbies: Reading

[0389] Values: Cherishing family

[0390] Lifestyle: Health-conscious

[0391] Emotion input example:

[0392] Tell us how you're feeling today:

[0393] joy

[0394] Using this prompt, the server can collect the user's characteristics and generate the AI ​​avatar's response through the emotion engine.

[0395] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0396] Step 1:

[0397] The user inputs the feature data

[0398] Users input their own characteristic data through the device's interface, which is designed to allow users to input detailed information such as hobbies, values, and lifestyle.

[0399] Input: User characteristic data (e.g., "Hobbies: Reading, Values: Cherishing family, Lifestyle: Health-conscious")

[0400] Output: Feature data input to the device

[0401] Step 2:

[0402] The device sends the input data to the server

[0403] The device sends the characteristic data entered by the user to the server. This data is sent using a secure communication protocol (e.g., HTTPS).

[0404] Input: Feature data entered into the device

[0405] Output: Feature data sent to the server

[0406] Step 3:

[0407] The server preprocesses the data

[0408] The server preprocesses the received feature data, specifically converting text data into numerical data and formatting it in a way that can be applied to machine learning models.

[0409] Input: Feature data sent to the server

[0410] Data processing: Convert text data into numerical data (e.g., "reading" into "hobby category: 1," "cherishing family" into "value category: 2," and "health-conscious" into "lifestyle category: 3")

[0411] Output: Preprocessed data

[0412] Step 4:

[0413] The server generates an AI avatar

[0414] The server uses the preprocessed data to generate an AI avatar, using a generative AI model to create an avatar that reflects the user's characteristics.

[0415] Input: Preprocessed data

[0416] Data calculation: Input quantified data into the generative AI model to generate an AI avatar

[0417] Output: Generated AI avatar

[0418] Step 5:

[0419] The user inputs emotion data

[0420] A user inputs his / her own emotional data through the terminal interface. The emotional data indicates information about emotions such as joy, sadness, and anger.

[0421] Input: User's emotional data (e.g., "How are you feeling today?: Happy")

[0422] Output: Emotion data input to the device

[0423] Step 6:

[0424] The device sends emotion data to the server.

[0425] The device transmits the emotion data entered by the user to the server, which also uses a secure communication protocol.

[0426] Input: Emotion data entered into the device

[0427] Output: Emotion data sent to the server

[0428] Step 7:

[0429] The server analyzes the emotional data and reflects it in the AI ​​avatar.

[0430] The server analyzes the received emotional data and uses an emotion engine to reflect it in the AI ​​avatar, specifically changing the AI ​​avatar's facial expressions and reactions based on the emotional data.

[0431] Input: Emotion data sent to the server

[0432] Data calculation: Analyzes emotional data using the emotion engine and sets the AI ​​avatar's reaction (e.g., analyzes the emotional data of "joy" and sets the AI ​​avatar to smile)

[0433] Output: AI avatar that responds according to emotions

[0434] (Application example 1)

[0435] Next, a description will be given of Application Example 1 of Embodiment 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."

[0436] While conventional systems could generate AI avatars that reflected the user's characteristics, they were unable to recognize the user's emotions and respond accordingly. Furthermore, they lacked the ability to recommend products based on the user's characteristics, limiting the shopping experience in virtual stores. This made it difficult to increase user satisfaction.

[0437] 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.

[0438] In this invention, the server includes means for generating an AI avatar that has learned the characteristics of each user, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for generating an AI avatar incorporating an emotion engine that recognizes the user's emotions and responds accordingly, means for recommending products based on the user's characteristics, and means for recognizing the user's emotions from camera footage.This makes it possible to respond according to the user's emotions and further recommend products based on the user's characteristics, thereby improving the shopping experience in virtual stores.

[0439] "User characteristics" refers to information specific to each individual user, such as the user's hobbies, values, and lifestyle.

[0440] An "AI avatar" is a virtual person or character created using artificial intelligence technology that reflects the user's characteristics.

[0441] An "emotion engine" is an artificial intelligence technology that recognizes the user's emotions and responds accordingly.

[0442] "Product recommendation" is the process of suggesting appropriate products based on the user's characteristics.

[0443] "Camera video" refers to video data acquired through a camera device.

[0444] "Matchmaking" is a process in which the generated AI avatars meet each other to determine their compatibility.

[0445] "Compatibility assessment" is the process of evaluating the compatibility between users based on the results of matchmaking.

[0446] The system for carrying out this invention consists of three main elements: a server, a terminal, and a user. The server includes means for generating an AI avatar that has learned the characteristics of each user, means for matching the generated AI avatars with each other, means for determining the compatibility between users based on the results of the matchmaking, means for generating an AI avatar incorporating an emotion engine that recognizes the user's emotions and responds accordingly, means for recommending products based on the user's characteristics, and means for recognizing the user's emotions from camera footage.

[0447] The terminal is a device such as a smartphone or head-mounted display, through which the user accesses the system. The terminal captures the user's image using a camera and sends it to the server. The server uses the captured image to perform emotion recognition and generates the reaction of the AI ​​avatar based on the results.

[0448] Specifically, the server processes and calculates data using software such as Python, OpenCV, Keras, and Scikit-learn. For example, it detects faces from camera footage and uses an emotion recognition model to recognize the user's emotions. Depending on the recognized emotion, the AI ​​avatar will respond by smiling or offering comforting words. It also converts the user's feature data into a TF-IDF vector and calculates the cosine similarity with a product database to recommend products.

[0449] As a concrete example, a user can access a virtual store using a smartphone, and an AI avatar can recommend an eco-bag to them. When the user smiles at the camera, the AI ​​avatar can respond by saying, "What a lovely smile!"

[0450] Examples of prompts for a generative AI model might include:

[0451] Input user characteristic data (hobbies, values, lifestyle, etc.). Next, input product database. Generate an AI avatar that recommends products based on the user's characteristics, recognizes the user's emotions from camera footage, and responds accordingly.

[0452] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0453] Step 1:

[0454] A user accesses the system through a terminal.

[0455] Input: User characteristic data (hobbies, values, lifestyle, etc.)

[0456] Output: User feature data is sent to the server

[0457] Specific operation: The user uses a smartphone or head-mounted display to access the system's interface and input their own characteristic data.

[0458] Step 2:

[0459] The server learns the user's characteristic data and generates an AI avatar.

[0460] Input: User feature data

[0461] Output: AI avatar that reflects the user's characteristics

[0462] Specific operation: The server uses Python to analyze the user's feature data and generates an AI avatar using a generative AI model.

[0463] Step 3:

[0464] The server acquires camera footage and performs emotion recognition.

[0465] Input: Camera image

[0466] Output: User emotion data

[0467] Specific operation: The device camera captures the user's image, detects the face using OpenCV, and recognizes the user's emotions using the Keras emotion recognition model.

[0468] Step 4:

[0469] The server uses an emotion engine to generate responses for the AI ​​avatar.

[0470] Input: User emotion data

[0471] Output: AI avatar response

[0472] Specific operation: Based on the recognized emotional data, the server generates a reaction from the AI ​​avatar, such as a smile or words of comfort.

[0473] Step 5:

[0474] The server recommends products using the user's feature data.

[0475] Input: User feature data, product database

[0476] Output: Recommended products list

[0477] Specific operation: The server uses Scikit-learn to convert the user's feature data into a TF-IDF vector, calculates the cosine similarity with the product database, and recommends products.

[0478] Step 6:

[0479] The device displays the AI ​​avatar's response and recommended products to the user.

[0480] Input: AI avatar response, recommended product list

[0481] Output: AI avatar responses and product list displayed to the user

[0482] Specific operation: The terminal displays the AI ​​avatar's reaction and the recommended product list sent from the server to the user, allowing the user to check and enjoy the shopping experience.

[0483] Example 2

[0484] Next, a description will be given of Example 2 of Form 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."

[0485] Conventional systems have difficulty accurately assessing the compatibility between users, and do not take into account the content or emotions of the conversation. Therefore, there is a need for a system that can more accurately assess the compatibility between users and efficiently advance the conversation.

[0486] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0487] In this invention, the server includes means for generating information processing devices that have learned the characteristics of individual users, means for having the generated information processing devices interact with each other, means for determining the compatibility between users based on the results of the interaction, means for analyzing the content of the interaction, the selection of topics, and the speed of reactions, and means for recognizing emotions shown by the information processing devices using an emotion recognition engine and evaluating the compatibility. This makes it possible to more accurately evaluate the compatibility between users and efficiently advance the interaction.

[0488] An "information processing device" is a device equipped with artificial intelligence to learn the characteristics of a user and engage in dialogue.

[0489] "Dialogue" is a process in which information processing devices converse with each other and exchange information with each other.

[0490] "Compatibility" is an index that indicates the compatibility between users, derived from the results of a dialogue between information processing devices.

[0491] The "contents of the conversation" refers to specific topics and words exchanged between information processing devices.

[0492] "Topic selection" refers to the theme or topic of conversation selected by the information processing device in a dialogue.

[0493] The "speed of response" refers to the time it takes for an information processing device to respond to a statement made by the other party in a conversation.

[0494] An "emotion recognition engine" is software or hardware for analyzing and recognizing emotions displayed by an information processing device.

[0495] "Emotion" refers to a psychological state such as joy, sadness, surprise, etc. that an information processing device shows during a conversation.

[0496] This invention is a system in which a server generates information processing devices (AI avatars) using a generative AI model and allows the generated information processing devices to interact with each other. A specific embodiment of this system is described below.

[0497] The server first generates an information processing device using a generative AI model (e.g., a natural language processing model). For example, a general natural language processing model is used as the generative AI model. The generated information processing device is set with individual personalities and topical tendencies.

[0498] Next, the server allows the generated information processing devices to interact with each other. The interaction takes the form of a conversation between the information processing devices. The server analyzes the content of the interaction, the choice of topics, and the speed of reactions in real time. This is done using natural language processing technology and machine learning algorithms.

[0499] Furthermore, the server recognizes emotions displayed by the information processing device using an emotion recognition engine. For example, general emotion analysis software is used for this emotion recognition engine. The server analyzes the emotions displayed by the information processing device using the emotion recognition engine and evaluates compatibility based on the analyzed emotions.

[0500] As a concrete example, the following prompt sentence is input to the generative AI model:

[0501] Example prompt sentence:

[0502] AI Avatar A: Hello, what would you like to talk about today?

[0503] AI Avatar B: Hello! I'd like to talk about a book I read recently.

[0504] AI Avatar A: That's great. What books have you read?

[0505] AI Avatar B: I recently read a book about the future of AI, and it was very interesting.

[0506] The server analyzes the flow of the conversation, the choice of topics, and the speed of reactions to calculate compatibility. It also uses an emotion recognition engine to recognize the emotions displayed by information processing devices A and B and reflect this in the compatibility evaluation. For example, if both devices show joy, the server determines that this is a good compatibility.

[0507] In this way, the server comprehensively evaluates the results of the dialogue between information processing devices, and the system more accurately judges the compatibility between users. This system enables users to efficiently find compatible partners.

[0508] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0509] Step 1:

[0510] The server generates an information processing device (AI avatar) using the generative AI model.

[0511] Input: A prompt for the generation AI model (e.g., "Generate a new AI avatar. It should have a cheerful personality and be interested in technology.")

[0512] Data processing: The generative AI model generates an AI avatar based on the prompt text and sets individual personality and topical tendencies.

[0513] Output: Generated AI avatar (information processing device)

[0514] Specific behavior:

[0515] The server inputs a prompt sentence into the generative AI model and generates an AI avatar, which is set with the specified personality and topical tendencies.

[0516] Step 2:

[0517] The server allows the generated information processing devices to interact with each other.

[0518] Input: Generated AI avatar (information processing device)

[0519] Data processing: The server initiates a dialogue between the AI ​​avatars and records the content of the conversation.

[0520] Output: Conversation log between AI avatars

[0521] Specific behavior:

[0522] The server pairs AI avatar A with AI avatar B and sends them an initial prompt. For example, the following conversation might take place:

[0523] AI Avatar A: Hello, what would you like to talk about today?

[0524] AI Avatar B: Hello! I'd like to talk about a book I read recently.

[0525] The server records this interaction.

[0526] Step 3:

[0527] The server analyzes the content of the conversation, the choice of topics, and the speed of responses.

[0528] Input: Conversation log between AI avatars

[0529] Data processing: Analyze the content of the conversation using natural language processing technology to extract topic selection and reaction speed.

[0530] Output: Analysis results (conversation topics, reaction times, etc.)

[0531] Specific behavior:

[0532] The server analyzes the interaction log and extracts data such as:

[0533] Conversation topics (e.g., technology, books)

[0534] Reaction time (e.g. AI avatar A's reaction time: 2 seconds, AI avatar B's reaction time: 3 seconds)

[0535] Step 4:

[0536] The server calculates the compatibility between AI avatars based on the analyzed data.

[0537] Input: Analysis results (conversation topics, reaction times, etc.)

[0538] Data processing: Compatibility is evaluated using statistical methods and machine learning algorithms.

[0539] Output: Compatibility score

[0540] Specific behavior:

[0541] The server calculates compatibility using the following algorithm:

[0542] Compatibility score = (match in conversation flow + match in topic selection + match in reaction speed) / 3

[0543] Step 5:

[0544] The server uses an emotion recognition engine to recognize emotions shown by the information processing device and evaluates compatibility.

[0545] Input: Conversation log between AI avatars

[0546] Data processing: Emotional data is extracted using an emotion recognition engine and reflected in compatibility evaluation.

[0547] Output: Emotion data, final compatibility evaluation

[0548] Specific behavior:

[0549] The server uses an emotion recognition engine to extract emotion data such as:

[0550] AI Avatar A's Emotion: Joy (80%)

[0551] AI Avatar B's Emotion: Joy (75%)

[0552] The server reflects this emotional data in the compatibility evaluation.

[0553] Step 6:

[0554] The server integrates the compatibility calculation results with the evaluation results of the emotion recognition engine and displays the final matchmaking results.

[0555] Input: Compatibility score, emotional data

[0556] Data processing: Compatibility scores and emotional data are integrated to make the final evaluation.

[0557] Output: Final matchmaking results

[0558] Specific behavior:

[0559] The server displays the result to the user as follows:

[0560] The compatibility score between AI Avatar A and AI Avatar B is 85. Both are expressing happiness, which indicates good compatibility.

[0561] (Application example 2)

[0562] Next, a description will be given of Application Example 2 of Form 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."

[0563] Conventional matchmaking systems using AI avatars have had the problem of being difficult to accurately assess compatibility between users. Furthermore, because compatibility is assessed without taking emotional reactions into account, it often fails to reflect compatibility in real-life communication. Furthermore, there was a lack of means to assess compatibility through conversation in a virtual space, making it difficult to provide users with a realistic experience.

[0564] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0565] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for recognizing the emotions of the AI ​​avatars using an emotion engine and evaluating the results of the matchmaking based on those emotions, means for the AI ​​avatars to converse with each other in a virtual space, and means for calculating compatibility by analyzing the content and progress of the conversation and emotional responses. This makes it possible to more accurately determine compatibility between users and provide a realistic communication experience.

[0566] An "AI avatar that has learned the characteristics of each individual user" is a virtual character created by artificial intelligence that learns the individual characteristics and preferences of each user and is based on that.

[0567] "Means for matching generated AI avatars" refers to a method for allowing generated AI avatars to interact with each other through conversation and determine compatibility based on the results.

[0568] "Method of determining compatibility between users based on matchmaking results" refers to a method of analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users based on that data.

[0569] An "emotion engine" is a technology that recognizes the emotions of an AI avatar and reacts and evaluates based on those emotions.

[0570] A "virtual space" is a computer-generated virtual environment in which users can engage in activities and interactions.

[0571] "Method of calculating compatibility by analyzing the content, progress, and emotional responses of conversations" refers to a method of analyzing in detail the content, progress, and emotional responses of conversations between AI avatars, and calculating compatibility based on this data.

[0572] The system for carrying out this invention consists of three main elements: a server, a terminal, and a user. The server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for recognizing the emotions of the AI ​​avatars using an emotion engine and evaluating the results of the matchmaking based on those emotions, means for the AI ​​avatars to converse with each other in a virtual space, and means for calculating compatibility by analyzing the content and progress of the conversation and emotional responses.

[0573] The server uses OpenAI's API to generate conversations between AI avatars. It uses EmotionEngine to recognize emotions in the conversation and evaluates compatibility based on the results. Specifically, the server generates AI avatars based on data provided by users, and the avatars converse with each other in a virtual space. The content and progress of the conversation, as well as emotional responses, are analyzed to calculate compatibility.

[0574] The terminal is a device such as a smartphone or head-mounted display, and provides the interface for users to use the system. Through the terminal, users can create their own AI avatar and conduct matchmaking with other users' AI avatars. The progress of the matchmaking can be visually confirmed on the terminal screen.

[0575] Users input their characteristics and preferences into the system and use an AI avatar generated based on that information. The results of the matchmaking are fed back to the user, allowing them to enjoy communicating in a virtual space with compatible partners before actually meeting them in person.

[0576] As a concrete example, consider a scenario in which User A and User B go on a blind date in a virtual store. User A's AI avatar asks, "Hello, how was your day today?", and User B's AI avatar replies, "It was a great day. How about you?" The emotions in this conversation are analyzed and a compatibility score is calculated.

[0577] An example of a prompt is as follows:

[0578] User A's AI avatar: Hello, how was your day?

[0579] User B's AI avatar: It was a great day. How about you?

[0580] User A's AI avatar: I had a good day too. What are you interested in these days?

[0581] User B's AI avatar: I've been interested in reading lately. What about you?

[0582] In this way, it becomes possible to more accurately judge compatibility between users and provide a realistic communication experience.

[0583] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0584] Step 1:

[0585] The server receives feature data provided by the user and generates an AI avatar that learns the features of each individual user. The input is the user's feature data, and the output is the generated AI avatar. The server uses a generative AI model based on this data to create the AI ​​avatar.

[0586] Step 2:

[0587] The server matches the generated AI avatars with each other. The input is the generated multiple AI avatars, and the output is the matchmaking results. The server sets up the AI ​​avatars to converse with each other in a virtual space and records the content of the conversation.

[0588] Step 3:

[0589] The server uses an emotion engine to recognize the emotions of the AI ​​avatar. The input is the conversation data during the matchmaking session, and the output is the recognized emotion data. The server uses the Emotion Engine to analyze the emotions for each message in the conversation.

[0590] Step 4:

[0591] The server analyzes the content and progress of the conversation, as well as emotional responses, to calculate compatibility. The input is conversation data and emotional data, and the output is a compatibility score. The server calculates compatibility based on this data and saves the results.

[0592] Step 5:

[0593] The terminal visually displays the matchmaking results to the user. The input is the compatibility score and detailed matchmaking data, and the output is the results screen displayed to the user. The terminal receives this data and displays it in a format that is easy for the user to understand.

[0594] Step 6:

[0595] Users can check the results of their matchmaking through their device and enjoy communicating with compatible partners in a virtual space. The input is the results screen displayed on the device, and the output is the user's next action. The user selects the next step based on the results.

[0596] In this way, the server, terminal, and user can work together to more accurately determine compatibility between users and provide a realistic communication experience.

[0597] Example 3

[0598] Next, a third embodiment of the third embodiment will be described. 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."

[0599] In conventional AI avatar systems, there are limited means for determining compatibility between users, and the lack of visually verifiable information makes it difficult for users to fully evaluate the compatibility between avatars and the process. Furthermore, there is a lack of means for visually providing feedback on the avatar's emotional reactions, making it difficult for users to see how their own emotions are reflected in the avatar.

[0600] The identification process by the identification processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means. In this invention, the server includes means for generating an AI avatar that has learned the characteristics of each user, means for matching the generated AI avatars with each other, means for determining the compatibility between users based on the results of the matchmaking, means for generating conversation content using a generative AI model, means for displaying the generated conversation content as text, means for generating emotional responses of the avatars using an emotion engine, and means for displaying the generated emotional responses as animation. This allows users to visually check the conversation content and emotional responses between the avatars, and allows them to evaluate the compatibility between the avatars and the process in more detail.

[0601] An "artificial intelligence avatar" is a virtual character that can learn the user's characteristics and engage in conversation and express emotions.

[0602] "Means for generating" refers to a method or device for creating an AI avatar, including the process of having it learn the characteristics of the user.

[0603] "Means for matching" refers to a method or device for allowing the generated AI avatars to converse with each other and communicate with each other.

[0604] "Means for determining compatibility" refers to a method or device for evaluating the compatibility between users based on the results of matchmaking between artificial intelligence avatars.

[0605] A "generative artificial intelligence model" refers to a machine learning model for natural language processing and dialogue generation, including, for example, a large-scale language model.

[0606] "Means for generating conversation content" refers to a method or device for generating conversation content between AI avatars using a generative AI model.

[0607] The "means for displaying as text" refers to a method or device for displaying the generated conversation content as character information on the user's terminal.

[0608] "Emotion engine" refers to software and algorithms for generating emotional responses of artificial intelligence avatars, including functions for emotion recognition and emotional expression.

[0609] "Means for generating an emotional response" refers to a method or apparatus for using an emotion engine to generate an emotional response of an artificial intelligence avatar.

[0610] The term "means for displaying as animation" refers to a method or device for displaying animation on a user's terminal to visually express the generated emotional reaction.

[0611] This invention relates to a matchmaking system for AI avatars, which allows users to visually check the content of conversations and emotional reactions between the avatars. Specific embodiments of this system are described below.

[0612] System configuration

[0613] This system consists of three main components: a server, a terminal, and a user. The server generates AI avatars, generates conversation content, and generates emotional responses. The terminal displays the generated conversation content and emotional responses. The user accesses the system and initiates matchmaking between avatars.

[0614] Hardware and software used

[0615] Server: A server with high-performance computing capabilities is required. Specifically, a server equipped with a GPU is preferable.

[0616] Generative AI model: A large-scale language model (such as GPT-4) is used as a generative AI model for performing natural language processing.

[0617] Emotion Engine: Uses emotion recognition software (e.g., Affectiva's emotion recognition software) to generate emotional responses for the avatar.

[0618] Device: The device used by the user can be anything that can run a web browser or dedicated app. Specifically, this includes PCs, smartphones, tablets, etc.

[0619] Program processing

[0620] The server first generates an AI avatar that learns the user's characteristics. To do this, it collects the user's input data and generates the avatar using a machine learning algorithm. The generated avatar reflects the user's characteristics and is capable of conversation and emotional expression.

[0621] Next, the server uses the generative AI model to generate conversation content between the avatars. It sends a prompt to the generative AI model and converts the returned conversation content into text. For example, the following prompt can be used:

[0622] "Please generate a conversation for a matchmaking session between two AI avatars. The conversation should be for Avatar A and Avatar B, and involve them introducing themselves to each other."

[0623] The generated conversation content is sent from the server to the device, which displays it on the screen, allowing the user to check the conversation between the avatars as text.

[0624] The server then uses an emotion engine to generate an emotional response for the avatar. The generated emotional response is sent to the device as an animation, which is then displayed on the screen. For example, if avatar A expresses joy, the response is displayed as an animated smile.

[0625] Specific examples

[0626] A user accesses the system and clicks the "Start Matchmaking" button. The server sends the following prompt to the generative AI model:

[0627] "Please generate a conversation for a matchmaking session between two AI avatars. The conversation should be for Avatar A and Avatar B, and involve them introducing themselves to each other."

[0628] The conversation content returned by the generative AI model is as follows:

[0629] Avatar A: "Hi, how was your day today?"

[0630] Avatar B: "Hi, I had a great day. How about you?"

[0631] The server sends the conversation content as text to the device, which then displays it on the screen. If Avatar A expresses joy using the emotion engine, the server generates a smiling animation to represent the avatar's reaction and sends it to the device. The device then displays this smiling animation on the screen.

[0632] Users can view the conversation content and animated smiles of the avatars displayed on the screen, and evaluate the compatibility between the AI ​​avatars and the process.

[0633] In this way, this system allows users to visually check the content of conversations and emotional reactions between avatars, and evaluate the compatibility between avatars and the process in more detail.The flow of the identification process in Example 3 will be explained using Figure 21.

[0634] Step 1:

[0635] The user accesses the system and clicks the "Start Matchmaking" button. The input is the user's operation, and the output is a request to start matchmaking sent to the server. Specifically, the user accesses the system through a web browser or a dedicated app and clicks the "Start Matchmaking" button on the screen.

[0636] Step 2:

[0637] The server generates an AI avatar that has learned the user's characteristics. The user's characteristic data is sent to the server as input, and the AI ​​avatar is generated as output. Specifically, the server uses a machine learning algorithm to learn the user's characteristics and generates an AI avatar based on that.

[0638] Step 3:

[0639] The server uses a generative AI model to generate conversation content between avatars. A prompt sentence is sent to the generative AI model as input, and the conversation content is generated as output. Specifically, the server sends the following prompt sentence to the generative AI model:

[0640] "Please generate a conversation for a matchmaking session between two AI avatars. The conversation should be for Avatar A and Avatar B, and involve them introducing themselves to each other."

[0641] Step 4:

[0642] The server sends the generated conversation content to the terminal as text. The generated conversation content is input, and text format data is sent to the terminal as output. In concrete terms, the server converts the generated conversation content into text format and sends it to the user's terminal.

[0643] Step 5:

[0644] The device displays the conversation content on the screen. The input is text data sent from the server, and the output is displayed on the screen. Specifically, the device displays the received text data on the screen. For example, avatar A's comments are displayed on the left side, and avatar B's comments are displayed on the right side.

[0645] Step 6:

[0646] The server uses an emotion engine to generate the avatar's emotional response. The generated conversation content is the input, and emotional response data is generated as the output. In concrete terms, the server uses the emotion engine to generate the avatar's emotional response based on the conversation content. For example, if avatar A shows joy, the corresponding emotion data is generated.

[0647] Step 7:

[0648] The server sends the generated emotional reactions as animations to the device. The input is the generated emotional reaction data, and the output is animation data sent to the device. Specifically, the server generates animations of the avatar's facial expressions and movements based on the emotional data, and sends them to the user's device.

[0649] Step 8:

[0650] The device displays the animation on the screen. The input is animation data sent from the server, and the output is displayed on the screen. In concrete terms, the device displays the received animation data on the screen. For example, an animation of avatar A smiling is played.

[0651] Step 9:

[0652] The user visually checks the conversation and emotional reactions between the AI ​​avatars. The input is the conversation content and animation displayed on the screen, and the output is the user's evaluation. Specifically, the user checks the conversation content and the avatars' emotional reactions displayed on the screen, and evaluates the compatibility between the AI ​​avatars and the process.

[0653] (Application example 3)

[0654] Next, a description will be given of Application Example 3 of Form Example 3. 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."

[0655] Conventional AI avatar systems have limited means for determining compatibility between users, and there is a lack of visually verifiable information. Even in customer service systems using AI avatars, improving customer satisfaction is an issue because the AI ​​avatars do not adequately introduce products to customers or answer their questions. Furthermore, there is no way for users to visually see how the AI ​​avatars compare and review products, making it difficult for users to evaluate the process.

[0656] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0657] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for the AI ​​avatars to introduce products to customers and answer their questions, means for visually checking how the AI ​​avatars compare and review products, and means for visually providing feedback on the emotional reactions of the AI ​​avatars.This allows users to visually check how the AI ​​avatars match with each other and compare and review products, and enables the server to efficiently introduce products to customers and answer their questions.

[0658] "User" refers to any individual who uses the system.

[0659] "AI avatar" refers to a virtual person or character generated using artificial intelligence technology.

[0660] "Matchmaking" refers to the process by which AI avatars converse and interact with each other.

[0661] "Compatibility" refers to the criteria for evaluating compatibility and suitability between users.

[0662] "Customer" refers to a person who wishes to purchase an item in a virtual store.

[0663] "Product introduction" refers to the act of an AI avatar explaining the features and benefits of a product to customers.

[0664] "Answering questions" refers to the act of an AI avatar providing appropriate information in response to a customer's question.

[0665] "Product comparison" refers to the process of comparing the features and benefits of multiple products.

[0666] A "review" refers to the act of rating or expressing opinions about a product.

[0667] "Visually identifiable" refers to a state in which a user can visually confirm information on a screen.

[0668] "Emotional responses" refer to emotions such as joy or sadness shown by an AI avatar.

[0669] "Feedback" refers to the process by which a system returns information to a user.

[0670] The system for implementing this invention consists of three main elements: a server, a terminal, and a user.

[0671] server

[0672] The server runs a program that has the following functions:

[0673] 1. AI avatar generation

[0674] The server generates an AI avatar that learns the characteristics of each individual user, using the user's profile information and past behavioral data. The generated AI avatar is a virtual character that reflects the user's characteristics.

[0675] 2. Matchmaking simulation

[0676] The server then matches the generated AI avatars with each other. In this process, the AI ​​avatars converse and interact with each other, and the server determines the compatibility between users based on the results.

[0677] 3. Visual Feedback

[0678] The server allows users to visually check the matchmaking process between the AI ​​avatars by displaying the content of the conversation as text and animating the avatars' facial expressions and movements.

[0679] 4. Emotion Engine

[0680] The server uses an emotion engine to generate the AI ​​avatar's emotional response and provides visual feedback. For example, if the AI ​​avatar expresses joy, the response is displayed as a smile on the screen.

[0681] 5. Product introduction and comparison

[0682] The server will have the function of allowing the AI ​​avatar to introduce products to customers and answer their questions, and will also allow users to visually see how the AI ​​avatars compare and review products with each other.

[0683] Terminal

[0684] A terminal is a device through which a user interacts with the system, including a smartphone, smart glasses, or a head-mounted display.

[0685] 1. Display function

[0686] The device displays visual feedback sent from the server, including the AI ​​avatar's speech, facial expressions, and movements.

[0687] 2. Input Function

[0688] The terminal accepts input from the user, including text input and voice input.

[0689] User

[0690] The user is an individual human who uses the system. The user interacts with an AI avatar through a terminal.

[0691] Specific examples

[0692] For example, if a user types "Tell me about this product," the server uses a generative AI model to generate an appropriate answer and provides feedback to the user through an AI avatar. Examples of prompt sentences include:

[0693] Generate an emotional response to the following text: Tell me about this product.

[0694] In this way, users can learn more about products through AI avatars, and can make better purchasing decisions by visually watching AI avatars compare and review products.

[0695] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0696] Step 1:

[0697] The server receives user profile information and past behavioral data as input and generates an AI avatar that learns the characteristics of each individual user. In this process, the server analyzes the data using a generative AI model and outputs a virtual character that reflects the user's characteristics.

[0698] Step 2:

[0699] The server matches the generated AI avatars with each other. In this step, it generates prompts for the AI ​​avatars to converse and interact with each other, and simulates a conversation based on these prompts. The input is the generated AI avatar and the prompt, and the output is a conversation log resulting from the matchmaking.

[0700] Step 3:

[0701] The server determines the compatibility between users based on the matchmaking results. It analyzes the conversation log as input and outputs a compatibility score. This analysis uses natural language processing technology to evaluate the emotions and tone of the conversation.

[0702] Step 4:

[0703] The server allows users to visually check the matchmaking process between the AI ​​avatars. Specifically, it displays the conversation content as text and animates the avatars' facial expressions and movements. The input is the conversation log, and the output is visual feedback.

[0704] Step 5:

[0705] The server uses an emotion engine to generate the AI ​​avatar's emotional response and provides visual feedback. For example, if the AI ​​avatar expresses joy, the response is displayed as a smile on the screen. The input is the conversation log and the prompt sentence of the emotion engine, and the output is an animation of the emotional response.

[0706] Step 6:

[0707] The server executes the function of an AI avatar to introduce products to customers and answer their questions. It receives questions from users as input, generates appropriate answers using a generative AI model, and provides feedback to the user through the AI ​​avatar. The output is detailed product information.

[0708] Step 7:

[0709] The server allows users to visually see how AI avatars compare and review products. It receives multiple product data as input, generates comparison results and reviews, and displays them visually. The output is the product comparison results and reviews.

[0710] Step 8:

[0711] The device displays visual feedback sent from the server, including the AI ​​avatar's conversation content, facial expressions, and movements. The input is the feedback data from the server, and the output is the visual display for the user.

[0712] Step 9:

[0713] The terminal accepts input from the user, which can be text or voice input. The input is the user's question or instruction, and the output is the request data to the server.

[0714] Step 10:

[0715] Users interact with the AI ​​avatar through their device, which allows them to obtain detailed product information, compare products with other AI avatars, and visually check reviews. The input is visual feedback from the device, and the output is the user's understanding and purchasing decision.

[0716] 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.

[0717] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ) 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.

[0718] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[0719] 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.

[0720] [Second embodiment]

[0721] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0722] 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.

[0723] 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).

[0724] 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.

[0725] 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.

[0726] 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).

[0727] 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.

[0728] 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.

[0729] 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.

[0730] 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.

[0731] 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.

[0732] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0733] "Example 1"

[0734] In one embodiment of the present invention, a system is provided that includes an interface through which a user can input their own characteristics. This interface is designed to allow detailed input of individual characteristics, such as the user's hobbies, values, and lifestyle. The input characteristics are used as learning data for an AI avatar, and each AI avatar reflects the user's characteristics.

[0735] "Example 2"

[0736] Next, the generated AI avatars are matched with each other. This matchmaking takes place in the form of a conversation between the AI ​​avatars, and compatibility is judged based on the content and progress of the conversation. Specifically, the flow of the conversation between the AI ​​avatars, choice of topics, speed of reaction, etc. are analyzed, and compatibility is calculated from these elements.

[0737] "Example 3"

[0738] Furthermore, users can visually check the progress of the matchmaking between the AI ​​avatars. Specifically, the conversation between the AI ​​avatars is displayed as text, and the avatars' facial expressions and movements are animated. This allows users to evaluate not only the compatibility between the AI ​​avatars, but also the process itself.

[0739] The processing flow of each embodiment will be described below.

[0740] "Example 1"

[0741] Step 1: The user accesses the system and enters their characteristics. Characteristics are items in which users can enter detailed information about their individual characteristics, such as hobbies, values, and lifestyle.

[0742] Step 2: The system generates an AI avatar based on the input characteristics. This AI avatar will reflect the user's characteristics.

[0743] "Example 2"

[0744] Step 1: The generated AI avatars begin meeting with each other. The meeting takes place in the form of a conversation between the AI ​​avatars.

[0745] Step 2: The system analyzes the flow of conversation between the AI ​​avatars, their choice of topics, and the speed of their reactions, and calculates compatibility based on these factors.

[0746] "Example 3"

[0747] Step 1: The user visually checks the matchmaking process between the AI ​​avatars through the system.

[0748] Step 2: Specifically, the content of conversations between AI avatars is displayed as text, and the avatars' facial expressions and movements are displayed as animations.

[0749] Step 3: This allows users to evaluate not only the compatibility between AI avatars, but also the process.

[0750] Example 1

[0751] Next, a description will be given of Example 1 of Form 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."

[0752] Conventional AI avatar generation systems have had difficulty efficiently generating AI avatars that reflect the characteristics of users and determining compatibility between users. Furthermore, there was a lack of a way to visually check the matchmaking process between the generated AI avatars, making it less convenient for users. Furthermore, there were also issues with the low reusability of AI avatars and poor time efficiency.

[0753] The identification process by the identification processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes a means having an interface for inputting the characteristics of individual users, a means for saving the input characteristics in a database, a means for creating a training dataset based on the saved data, a means for generating an AI avatar using a generative AI model, a means for matching the generated AI avatars with each other, and a means for determining the compatibility between users based on the results of the matchmaking. This makes it possible to efficiently generate AI avatars that reflect the characteristics of users and accurately determine the compatibility between users. In addition, the matchmaking process between the generated AI avatars can be visually confirmed, improving convenience for users. Furthermore, the reusability of AI avatars is increased, and time efficiency is improved.

[0754] "User" refers to an individual who uses the system to input their own characteristics and generate an AI avatar.

[0755] "Interface" refers to the means by which a screen or form is presented to a user to input their characteristics.

[0756] "Characteristics" refers to individual attributes and information such as a user's hobbies, values, and lifestyle.

[0757] "Database" refers to an information system for storing and managing input characteristics.

[0758] A "training dataset" refers to a collection of data used to train an AI model based on stored feature data.

[0759] A "generative AI model" refers to an algorithm or program for generating an AI avatar based on a training dataset.

[0760] An "AI avatar" refers to a virtual character generated to reflect the user's characteristics.

[0761] "Matchmaking" refers to the process by which generated AI avatars interact with each other and determine their compatibility.

[0762] "Compatibility" refers to an indicator that shows how closely users' characteristics and values ​​match.

[0763] "Means for visual confirmation" refers to a function that allows users to view the matchmaking process between the generated AI avatars on the screen.

[0764] "Reusability" refers to the property that an AI avatar that has been trained can be used any number of times.

[0765] "Time efficiency" refers to the speed at which the AI ​​avatar generation and matchmaking process is carried out.

[0766] The present invention provides a system having an interface for allowing a user to input his / her own characteristics. A specific embodiment of this system will be described below.

[0767] First, the user enters their characteristics using an interface on a web browser. This interface is built using HTML, CSS, and JavaScript. The user can enter detailed information about their individual characteristics, such as hobbies, values, and lifestyle. For example, the user might enter "Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious."

[0768] The terminal then sends the data entered by the user to the server using an HTTP POST request. This communication uses Axios, a JavaScript HTTP client library. The terminal sends the entered data to the server in JSON format.

[0769] The server analyzes the received data and saves it in a database. A relational database such as MySQL or PostgreSQL is used for the database. The server analyzes the received data and saves it in the database as "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-Oriented."

[0770] Next, the server creates a training dataset based on the stored data. This dataset reflects the user's characteristics. The server creates a training dataset for the AI ​​model based on the data retrieved from the database.

[0771] The server uses the generated dataset to generate an AI avatar. This process uses a generative AI model (such as GPT-3 or BERT). The server uses the generative AI model to generate an AI avatar with the following characteristics: Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious.

[0772] Finally, the device displays the AI ​​avatar received from the server. A front-end framework such as React.js is used for display. The device displays the AI ​​avatar received from the server using React.js, showing the user, "Hello, I'm a health-conscious avatar who loves reading and values ​​family."

[0773] As a specific example, if a user enters "Hobbies: Reading, Values: Valuing family, Lifestyle: Health-conscious," the server will generate an AI avatar based on this data and display to the user, "Hello, I'm a health-conscious avatar who loves reading and values ​​family."

[0774] An example prompt is, "Based on the user's hobbies, values, and lifestyle, please generate an AI avatar that reflects their individual characteristics. For example, if their hobby is reading, their values ​​are family-oriented, and their lifestyle is health-conscious, please create an AI avatar with those characteristics."

[0775] In this way, the system can efficiently generate AI avatars that reflect the characteristics of each user and determine compatibility between users. Furthermore, users can visually check the matchmaking process between the generated AI avatars, improving convenience for users. Furthermore, the reusability of AI avatars is enhanced, improving time efficiency.

[0776] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0777] Step 1:

[0778] User enters characteristics

[0779] A user inputs their own characteristics using an interface on a web browser. This interface is built using HTML, CSS, and JavaScript. Specifically, the user inputs "Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" into the form and clicks the submit button. The input data includes individual characteristics such as hobbies, values, and lifestyle.

[0780] Step 2:

[0781] The device sends the data to the server

[0782] The device sends the data entered by the user to the server using an HTTP POST request. This communication uses Axios, a JavaScript HTTP client library. Specifically, the device sends the data "Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" in JSON format to the server. The input is the feature data entered by the user, and the output is the JSON data sent to the server.

[0783] Step 3:

[0784] The server receives and stores the data

[0785] The server analyzes the received data and stores it in a database. A relational database such as MySQL or PostgreSQL is used for the database. Specifically, the server analyzes the received data and stores it in the database as "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-Oriented." The input is JSON data sent from the device, and the output is the user feature data stored in the database.

[0786] Step 4:

[0787] The server creates the training dataset.

[0788] The server creates a training dataset based on the stored data. This dataset reflects the user's characteristics. Specifically, the server retrieves the following data from the database: "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" and creates a training dataset for the AI ​​model. The input is the user characteristic data retrieved from the database, and the output is the training dataset.

[0789] Step 5:

[0790] The server generates an AI avatar

[0791] The server uses the generated dataset to generate an AI avatar. This process uses a generative AI model (such as GPT-3 or BERT). Specifically, the server uses the generative AI model to generate an AI avatar with the following characteristics: Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious. The input is the training dataset, and the output is the generated AI avatar.

[0792] Step 6:

[0793] The device displays an AI avatar

[0794] The device displays the AI ​​avatar received from the server. A front-end framework such as React.js is used for display. Specifically, the device displays the AI ​​avatar received from the server using React.js, showing the user "Hello, I'm a health-conscious avatar who loves reading and values ​​family." The input is the AI ​​avatar received from the server, and the output is the AI ​​avatar displayed to the user.

[0795] (Application example 1)

[0796] Next, a description will be given of Application Example 1 of Form 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."

[0797] Conventional content distribution services have the problem of not providing sufficient personalized content recommendations based on users' hobbies, values, and lifestyles. Furthermore, there is a lack of systems that allow users to input their own detailed characteristics and then recommend the most appropriate content based on that information, which can lead to a decline in user satisfaction.

[0798] 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.

[0799] In this invention, the server includes a means for generating an AI avatar that has learned the characteristics of each user, a means for recommending optimal content to the user based on the generated AI avatar, and a means for providing the recommended content to the user. This enables personalized content recommendations based on the user's characteristics, thereby improving user satisfaction.

[0800] "User" refers to any individual who uses the system.

[0801] "Characteristics" refers to individual attributes such as a user's hobbies, values, and lifestyle.

[0802] An "AI avatar" refers to an artificial intelligence model that learns the user's characteristics and is generated based on them.

[0803] "Content" refers to information and entertainment such as movies, music, and articles.

[0804] "Recommendation" refers to the act of selecting and presenting the most suitable content based on the user's characteristics.

[0805] "Providing" refers to the act of actually showing or making available the recommended content to a user.

[0806] "Visual confirmation means" refers to a method that allows users to visually confirm the recommendation process and results through a screen or other means.

[0807] "Time-efficient" refers to a state in which a goal can be achieved effectively in a short amount of time.

[0808] The system for implementing this invention comprises a server and a user terminal. The server includes a means for generating an AI avatar that has learned the characteristics of each user, a means for recommending optimal content to the user based on the generated AI avatar, and a means for providing the recommended content to the user.

[0809] Program processing explanation

[0810] 1. User characteristics input

[0811] The user terminal provides an interface for users to input their own characteristics (hobbies, values, lifestyle, etc.). This interface is implemented as a smartphone application. The characteristics input by the user are sent to the server.

[0812] 2. AI avatar generation

[0813] The server generates an AI avatar based on the received user feature data. This AI avatar is trained using a generative AI model. Specifically, it uses Python and the scikit-learn library to vectorize the user's features and save them as an AI avatar.

[0814] 3. Content Recommendations

[0815] The server then recommends the most suitable content to the user based on the generated AI avatar. The recommendation algorithm is implemented using TF-IDF vectorization and cosine similarity calculation, which selects the content that most closely matches the user's characteristics.

[0816] 4. Provision of Content

[0817] The server transmits the recommended content to the user terminal for visual confirmation by the user, and the user terminal provides an interface for displaying the recommended content.

[0818] Specific examples

[0819] For example, if a user enters the following characteristics:

[0820] Hobbies: Watching movies, listening to music

[0821] Values: Cherishing time with family

[0822] Lifestyle: Health-conscious

[0823] In this case, an example prompt to input to the generative AI model would be:

[0824] The user's hobbies are watching movies and listening to music, and they value spending time with their family. Their lifestyle is health-conscious. Please recommend the most suitable content for this user.

[0825] Based on this prompt, the server will recommend the most suitable content for the user, such as movies, music, articles, etc., and provide it to the user's device, allowing the user to enjoy personalized content based on their own characteristics.

[0826] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0827] Step 1:

[0828] The user device provides an interface where the user can input their own characteristics (hobbies, values, lifestyle, etc.). The characteristics input by the user are sent to the server in JSON format. Examples of input data include "Hobbies: watching movies, listening to music," "Values: valuing time with family," and "Lifestyle: health-conscious."

[0829] Step 2:

[0830] The server analyzes the received user feature data and generates an AI avatar using a generative AI model. Specifically, it uses Python and the scikit-learn library to vectorize the user features using TF-IDF and save them as an AI avatar. The input data are the user features, and the output data is the vectorized AI avatar.

[0831] Step 3:

[0832] The server recommends the most suitable content to the user based on the generated AI avatar. The recommendation algorithm is implemented using TF-IDF vectorization and cosine similarity calculation. Specifically, it compares the user's feature vector with the vector in the content list and selects the content with the highest similarity. The input data is the AI ​​avatar and the content list, and the output data is the recommended content.

[0833] Step 4:

[0834] The server transmits the recommended content to the user terminal. The user terminal provides an interface for displaying the recommended content. Specifically, the recommended content, such as movies, music, or articles, is displayed on the user's screen. The input data is the recommended content, and the output data is the content displayed on the user terminal.

[0835] Step 5:

[0836] The user can visually check and use the recommended content. When the user selects content, the selection information is fed back to the server and used for the next recommendation. The input data is the user's selection information, and the output data is the feedback information that will be reflected in the next recommendation.

[0837] Example 2

[0838] Next, a description will be given of Example 2 of Form 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."

[0839] Conventional matchmaking systems using AI avatars have had difficulty accurately determining compatibility between users. Furthermore, they lacked the functionality to analyze the content and progress of conversations in real time and calculate compatibility, making it difficult for users to obtain satisfactory results. Furthermore, there was no way to visually check the matchmaking process between the generated AI avatars, making it difficult for users to understand or accept the system.

[0840] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0841] In this invention, the server includes a means for a user to input a prompt sentence, a means for generating an AI avatar using a generative AI model, a means for matching the generated AI avatars, a means for analyzing the content and progress of the conversation, a means for calculating the compatibility between the AI ​​avatars, a means for notifying the user of the compatibility results, a means for the user to visually confirm the progress of the matchmaking between the AI ​​avatars, and a means for allowing the AI ​​avatar to be used repeatedly once trained, thereby allowing the matchmaking to proceed time-efficiently. This makes it possible to accurately determine the compatibility between users and provide satisfactory results. Furthermore, since users can visually confirm the progress of the matchmaking, understanding and satisfaction are improved.

[0842] A "prompt sentence" is an instruction sentence that a user inputs to a generative AI model to generate an AI avatar.

[0843] A "generative AI model" is an artificial intelligence model for generating an AI avatar based on an input prompt sentence.

[0844] An "AI avatar" is a virtual character generated by a generative AI model with specific characteristics and personality.

[0845] "Matchmaking" is a process in which generated AI avatars converse with each other to determine their compatibility.

[0846] "Means for analyzing the content and progress of conversations" refers to a function that monitors and evaluates in real time the flow of conversations between AI avatars, topic selection, speed of reactions, etc.

[0847] The "means for calculating compatibility" is a function that calculates the compatibility between AI avatars in the form of numbers, graphs, etc. based on the results of conversation analysis.

[0848] The "means for notifying compatibility results" is a function for notifying the user of the calculated compatibility results.

[0849] "Visual confirmation means" is a function that allows users to visually check the progress of matchmaking between AI avatars.

[0850] "A time-efficient means of conducting matchmaking" is a function that allows the AI ​​avatar to progress through matchmaking in a short amount of time, and can be used any number of times once it has been trained.

[0851] This invention is a system in which a user inputs a prompt sentence, generates an AI avatar using a generative AI model, and matches the generated AI avatars with each other. Specific embodiments of this system are described below.

[0852] First, the user inputs a prompt using a terminal. The prompt is an instruction for the generative AI model to generate an AI avatar, such as "You are a 30-year-old man whose hobbies are reading and hiking. Please introduce yourself."

[0853] The server then receives the prompt and generates an AI avatar using a generative AI model (e.g., OpenAI's GPT-4 or Google's BERT). The generated AI avatar has information such as a self-introduction and hobbies based on the prompt.

[0854] The server pairs the generated AI avatars and initiates matchmaking. Matchmaking takes place by having the AI ​​avatars converse with each other. The server monitors the content and progress of the conversation in real time and analyzes the following factors:

[0855] Conversation flow: Check that the conversation is flowing smoothly, i.e., that the topic is kept on topic and that the conversation is unfolding naturally.

[0856] Topic selection: Evaluate whether the topics chosen by the AI ​​avatar are appropriate, for example, ensuring that topics are based on shared interests.

[0857] Reaction speed: Measures how quickly the AI ​​avatar responds to what the other person says. Evaluates whether it responds quickly and appropriately.

[0858] The server analyzes these factors comprehensively and calculates the compatibility between the AI ​​avatars. The compatibility is displayed in the form of numbers and graphs. For example, if the compatibility is evaluated as 80%, the server notifies the user of the result.

[0859] Furthermore, the server provides a means for users to visually check the matchmaking process between the AI ​​avatars, allowing them to intuitively understand the progress and results of the matchmaking.

[0860] Furthermore, once the AI ​​avatar has been trained, it can be used any number of times, allowing for time-efficient matchmaking, enabling users to conduct multiple matchmaking sessions in a short amount of time and find the perfect partner.

[0861] The above is a specific embodiment for carrying out the present invention.

[0862] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0863] Step 1:

[0864] The user enters a prompt statement.

[0865] The user uses a terminal to input a prompt to the generative AI model to generate an AI avatar. For example, the prompt might be, "You are a 30-year-old man whose hobbies are reading and hiking. Please introduce yourself." The input prompt is then sent to the server.

[0866] Step 2:

[0867] The server generates an AI avatar using the generative AI model.

[0868] The server receives the prompt sent by the user and inputs it into a generative AI model (e.g., OpenAI's GPT-4). The generative AI model generates an AI avatar based on the prompt and determines the avatar's characteristics and personality. The data of the generated AI avatar is stored on the server.

[0869] Step 3:

[0870] The server matches the generated AI avatars with each other.

[0871] The server pairs the generated AI avatars and initiates matchmaking. Matchmaking takes place in the form of conversation between the AI ​​avatars. The server instructs each AI avatar to start a conversation and collects conversation logs in real time.

[0872] Step 4:

[0873] The server analyzes the content and progress of the conversation.

[0874] The server analyzes the collected conversation logs and evaluates the following factors:

[0875] Conversation flow: Check that the conversation is flowing smoothly, i.e., that the topic is kept on topic and that the conversation is unfolding naturally.

[0876] Topic selection: Evaluate whether the topics chosen by the AI ​​avatar are appropriate, for example, ensuring that topics are based on shared interests.

[0877] Reaction speed: Measures how quickly the AI ​​avatar responds to what the other person says. Evaluates whether it responds quickly and appropriately.

[0878] Step 5:

[0879] The server calculates the compatibility between AI avatars.

[0880] The server calculates the compatibility between the AI ​​avatars based on the data collected in step 4. The compatibility is displayed in the form of a number or a graph. For example, if the compatibility is evaluated as 80%, the server generates the result.

[0881] Step 6:

[0882] The server notifies the user of the compatibility results.

[0883] The server notifies the user of the calculated compatibility results, which the user can check via their device. This allows the user to decide their next action based on the results of the matchmaking.

[0884] Step 7:

[0885] To provide a means for a server to visually check the state of a matchmaking session.

[0886] The server provides an interface that allows users to visually check the matchmaking process between AI avatars, and users can intuitively understand the progress and results of the matchmaking process through their devices.

[0887] Step 8:

[0888] The server makes the AI ​​avatars reusable.

[0889] The server allows the generated AI avatar to be trained once and then used again and again, allowing users to conduct multiple matchmaking sessions in a short amount of time and find the perfect partner.

[0890] (Application example 2)

[0891] Next, a description will be given of Application Example 2 of Form 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."

[0892] Conventional matchmaking systems using AI avatars have had difficulty accurately determining compatibility between users. They also lack the ability to take into account detailed factors such as the flow of conversation, topic selection, and speed of response, resulting in insufficient calculations of compatibility. Furthermore, there was a lack of a way for users to visually check the progress of the matchmaking, which hindered the user experience.

[0893] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining the compatibility between users based on the results of the matchmaking, means for analyzing the flow of conversation, topic selection, and reaction speed, and means for calculating compatibility. This makes it possible to more accurately determine the compatibility between users and visually check the matchmaking process.

[0894] "Means for generating an AI avatar that has learned the characteristics of each individual user" refers to a method or device for learning the individual characteristics and preferences of each user and generating an AI avatar based on that.

[0895] "Means for matching generated AI avatars" refers to a method or device for matching generated AI avatars through conversation.

[0896] "Means for determining compatibility between users based on matchmaking results" refers to a method or device for analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users based on the results.

[0897] "Means for analyzing conversation flow, topic selection, and speed of response" refers to methods and devices for analyzing the flow of conversations between AI avatars, the topics selected, the speed of response, etc.

[0898] The "means for calculating compatibility" refers to a method or device for quantifying the compatibility between users based on the results of conversation analysis.

[0899] "Means for visual confirmation" refers to a method or device that allows users to visually confirm the matchmaking process between AI avatars.

[0900] "A means for efficiently conducting matchmaking in a time-efficient manner" refers to a method or device that allows an AI avatar to be trained once and then used any number of times, allowing matchmaking to proceed efficiently in a short amount of time.

[0901] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The following explains how each element functions.

[0902] First, the server has a means for generating an AI avatar that has learned the characteristics of each individual user. This means generates an AI avatar using a generative AI model based on the user's input data (e.g., profile information and past conversation history). The generated AI avatar reflects the user's characteristics and has individual preferences and personality.

[0903] Next, the server has a means for matching the generated AI avatars with each other. This means generates prompts for the generated AI avatars to converse with each other, and simulates the conversation using the generative AI model. For example, the following prompts are used:

[0904] Avatar 1: Hello, how are you doing these days?

[0905] Avatar 2: I'm fine. How about you?

[0906] Start a conversation.

[0907] The server has a means for analyzing the flow of conversation, topic selection, and speed of response. This means analyzes the generated conversation text and evaluates elements such as the flow of conversation, topic selection, and speed of response. This allows data to be obtained for quantifying the quality of the conversation and compatibility.

[0908] Furthermore, the server has a means for calculating compatibility, which quantifies the compatibility between users based on the results of conversation analysis. For example, if the conversation flows smoothly and the choice of topics reflects common interests, a high compatibility score is calculated.

[0909] The device has a means for users to visually check the matchmaking process. This allows users to visually check the conversation between the AI ​​avatars in real time. For example, they can use a smartphone or head-mounted display to visually experience the conversation in a virtual space.

[0910] Users can control the AI ​​avatar through their device and check the results of their matchmaking. Users can customize their own avatar and check compatibility by talking with other users' avatars.

[0911] This system allows users to more accurately judge compatibility between each other and visually check the matchmaking process. Specifically, it provides a realistic conversation experience by simulating conversations using a generative AI model and generating conversations using prompts.

[0912] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0913] Step 1:

[0914] The server receives input data from the user. The input data includes the user's profile information and past conversation history. Based on this data, a generative AI model is used to generate an AI avatar that learns the user's characteristics. The generated AI avatar reflects the user's individual preferences and personality. The input is the user's data, and the output is the generated AI avatar.

[0915] Step 2:

[0916] The server generates a prompt to match the generated AI avatars. The prompt is text to start a conversation between the AI ​​avatars. For example, it takes the form "Avatar 1: Hello, how are you?\nAvatar 2: I'm fine. How about you?\nLet's start the conversation." The input is the generated AI avatar, and the output is the prompt.

[0917] Step 3:

[0918] The server uses a generative AI model to simulate a conversation based on the prompt sentence. The generative AI model receives the prompt sentence as input and generates a conversation between AI avatars. The generated conversation is output in text format. The input is the prompt sentence, and the output is the generated conversation text.

[0919] Step 4:

[0920] The server analyzes the generated conversation text. The analysis includes elements such as the flow of the conversation, topic selection, and speed of response. These elements are evaluated to obtain data to quantify the quality of the conversation and compatibility. The input is the generated conversation text, and the output is the analysis results.

[0921] Step 5:

[0922] The server calculates the compatibility between users based on the analysis results. Compatibility is high when the conversation flows smoothly and the topic choices reflect common interests. Compatibility is quantified and fed back to the user. The input is the analysis results, and the output is a compatibility score.

[0923] Step 6:

[0924] The terminal allows the user to visually check the progress of the matchmaking. The terminal visually displays the generated conversation text, allowing the user to check the conversation in real time. The input is the generated conversation text, and the output is the visual display.

[0925] Step 7:

[0926] Users control the AI ​​avatar through their device and check the matchmaking results. Users can customize their own avatar and check compatibility by conversing with other users' avatars. The input is a compatibility score, and the output is user feedback.

[0927] Example 3

[0928] Next, a description will be given of Example 3 of Form Example 3. 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."

[0929] In conventional systems using AI avatars, there were limited means for determining compatibility between users, and the system was unable to visually confirm it. It was also difficult to evaluate the process of dialogue between AI avatars, making it less convenient for users. Furthermore, generating AI avatars and simulating dialogue took time, making it difficult to proceed efficiently.

[0930] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[0931] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for having the generated AI avatars converse with each other, means for determining compatibility between users based on the results of the conversation, means for the user to input a prompt sentence, means for simulating a conversation between the generated AI avatars, means for generating animation data of the conversation content and the avatars, means for transmitting the generated data to a terminal, means for the terminal to display the conversation content and avatar animation, and means for the user to evaluate the compatibility between the AI ​​avatars and the progress of the conversation. This allows the user to visually check the state of the conversation between the AI ​​avatars and evaluate the process. Furthermore, once the AI ​​avatars have been trained, they can be used repeatedly, allowing for time-efficient conversations.

[0932] An "artificial intelligence avatar" is a virtual character that learns the user's characteristics and is generated to simulate conversations and actions.

[0933] A "prompt" is a text of instructions or questions entered by a user, which serves as the basis for generating an AI avatar and simulating a dialogue.

[0934] "Dialogue" refers to conversations and communication between artificial intelligence avatars, and is simulated based on user input.

[0935] "Simulation" is the process of virtually recreating the interactions and behaviors of artificially intelligent avatars, and is done using a generative AI model.

[0936] "Animation data" is data that visually expresses the facial expressions and movements of an artificial intelligence avatar, and is displayed to the user in real time.

[0937] A "terminal" is a device that allows a user to input prompts and visually check the dialogue and animation of an artificial intelligence avatar.

[0938] A "generative AI model" is a machine learning model used to generate artificial intelligence avatars and simulate dialogue, and includes, for example, natural language processing models.

[0939] "Compatibility" refers to the suitability and compatibility between users, as determined based on the results of conversations between AI avatars.

[0940] "Evaluation" refers to the judgment or feedback that the user gives based on the process and results of the dialogue between AI avatars.

[0941] The present invention is a system that uses an artificial intelligence avatar and provides a means for determining compatibility between users. A specific embodiment of this system will be described below.

[0942] First, the user inputs a prompt using the terminal. The prompt specifies the theme and content of the conversation between the AI ​​avatars. For example, the user can input a prompt such as "Start a matchmaking session between AI avatars discussing hobbies."

[0943] The server then receives the prompt and generates an AI avatar using a generative AI model, such as the natural language processing model GPT-4. The generated AI avatar is configured to have different personalities and hobbies based on the prompt.

[0944] The server simulates a conversation between the generated AI avatars. The content of the simulated conversation is generated in text format. For example, Avatar A might say, "I like reading. I especially like mystery novels," and Avatar B might reply, "I like reading, too. I've been hooked on historical novels lately."

[0945] Furthermore, the server generates animation data for the avatars' facial expressions and movements based on the content of the conversation. For example, it generates an animation of avatar A smiling and talking, and avatar B nodding.

[0946] The generated text data and animation data are sent from the server to the device. Based on the received data, the device displays the dialogue content as text to the user and animates the avatars' facial expressions and movements. This allows the user to visually confirm the dialogue between the AI ​​avatars.

[0947] Finally, the user can evaluate the compatibility and dialogue process between the AI ​​avatars based on the dialogue content and animation displayed on the device. For example, they can evaluate that "Avatar A and Avatar B seem to get along well."

[0948] This system allows users to visually check the state of conversations between AI avatars and evaluate the process. Furthermore, once the AI ​​avatars have been trained, they can be used any number of times, allowing for time-efficient conversations.

[0949] As an example, the following prompt sentence is used:

[0950] "Starting matchmaking with AI avatars that discuss hobbies"

[0951] "Starting a matchmaking session with AI avatars to discuss travel"

[0952] "We're starting a matchmaking session for AI avatars that discuss music."

[0953] The above is an embodiment of the present invention. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0954] Step 1:

[0955] The user enters a prompt statement.

[0956] The user enters a prompt into the input field of the device, for example, "Start a matchmaking session for an AI avatar to discuss hobbies." The entered prompt is sent from the device to the server.

[0957] Step 2:

[0958] The server receives the prompt sentence and generates an artificial intelligence avatar using a generative AI model.

[0959] The server receives a prompt sent by the user. It then uses a generative AI model (e.g., a natural language processing model) to generate two AI avatars based on the prompt. The generated avatars are configured to have different personalities and hobbies. The input is the prompt, and the output is the generated AI avatars.

[0960] Step 3:

[0961] The server simulates conversations between artificial intelligence avatars.

[0962] The server simulates a conversation between the generated AI avatars. Using the generative AI model, Avatar A and Avatar B have a conversation based on a prompt. For example, Avatar A says, "I like reading," and Avatar B replies, "I like reading, too." The input is the generated AI avatar, and the output is the simulated conversation content.

[0963] Step 4:

[0964] The server generates the conversation content and avatar animation data.

[0965] The server generates the simulated conversation content in text format. It also generates the avatars' facial expressions and movements as animation data. For example, it generates an animation of avatar A smiling and talking, and avatar B nodding. The input is the simulated conversation content, and the output is text data and animation data.

[0966] Step 5:

[0967] The server transmits the generated data to the terminal.

[0968] The server sends the generated text data and animation data of the conversation to the terminal. The input is the text data and animation data, and the output is sent to the terminal.

[0969] Step 6:

[0970] The device displays the conversation and an animated avatar.

[0971] The device displays the conversation content as text to the user based on the data received from the server. It also displays the avatar's facial expressions and movements as animation. For example, avatar A and avatar B are displayed on the screen, and the conversation content and facial expressions of each are displayed in real time. The input is the data received from the server, and the output is what is displayed to the user.

[0972] Step 7:

[0973] Users evaluate the compatibility between AI avatars and the conversation process.

[0974] The user checks the conversation content and animation displayed on the device and evaluates the compatibility between the AI ​​avatars and the progress of the conversation. For example, the user may evaluate that "Avatar A and Avatar B seem to get along well." The input is the displayed conversation content and animation, and the output is the user's evaluation.

[0975] (Application example 3)

[0976] Next, a description will be given of Application Example 3 of Form Example 3. 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."

[0977] Conventional AI avatar systems have limited means for determining compatibility between users, and users cannot visually check the interactions between AI avatars. Furthermore, when introducing or comparing products, they lack the interactive functionality to allow users to ask questions in real time and receive answers. This has led to problems such as reduced user satisfaction and convenience.

[0978] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[0979] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for the AI ​​avatars to introduce and compare products with each other, means for visually checking this situation, and means for the user to ask the AI ​​avatar questions and receive responses in real time, thereby enabling users to interactively obtain product information while visually checking the interactions between the AI ​​avatars.

[0980] An "AI avatar" is a virtual character that learns the characteristics of each individual user and is generated based on those characteristics.

[0981] "Matchmaking" is a process in which generated AI avatars converse and interact with each other, and their compatibility is determined based on the results.

[0982] The "means of determining compatibility" is a method for analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users.

[0983] "Means for introducing and comparing products" is a function that allows an AI avatar to provide information about specific products and compare multiple products.

[0984] "Visual confirmation" refers to a method that allows users to visually observe the interactions and animations between AI avatars.

[0985] "Real-time reception" is a feature that allows users to ask questions to an AI avatar and receive answers instantly.

[0986] A system for implementing the present invention includes a server, a user terminal, and an AI avatar generation program. Specific embodiments are described below.

[0987] System Configuration

[0988] 1. Server:

[0989] Hardware: A server equipped with a high-performance CPU, GPU, and memory.

[0990] Software: TensorFlow (generative AI model), Unity (animation display), OpenCV (image processing).

[0991] 2. User Device:

[0992] Hardware: Smartphone, Head-Mounted Display (HMD).

[0993] Software: Dedicated application (Unity-based).

[0994] Program processing

[0995] 1. AI Avatar Generation:

[0996] The server uses TensorFlow to build a generative AI model and have it learn the user's characteristics.

[0997] When a user enters a prompt sentence, the generative AI model generates an AI avatar based on its features.

[0998] 2. Conducting matchmaking:

[0999] The server matches the generated AI avatars with each other.

[1000] Based on the results of the matchmaking, the server determines the compatibility between users.

[1001] 3. Product introduction and comparison:

[1002] The server generates scenarios in which AI avatars introduce and compare products.

[1003] The user's device uses Unity to display the AI ​​avatar's facial expressions and movements in animated form.

[1004] 4. Visual Verification:

[1005] The user device uses OpenCV to track the user's gaze and facial expressions in real time, providing an interactive experience.

[1006] Users can visually see the interactions between AI avatars.

[1007] 5. Real-time Questions and Answers:

[1008] Users can ask questions to the AI ​​avatar.

[1009] The server generates an answer to the question in real time and transmits it to the user terminal.

[1010] Specific examples

[1011] The user puts on a smartphone or HMD and accesses a virtual store. An AI avatar appears on the screen, and the user enters a prompt: "Tell me about the features of my new smartphone." The AI ​​avatar responds to the question by comparing the features of several smartphones. The conversation is displayed as text, and the avatar's facial expressions and movements are animated, allowing the user to visually confirm the conversation.

[1012] Prompt Sentence Examples

[1013] "What's the difference between the latest smartphones and tablets?"

[1014] "Write a review for this product"

[1015] How does this product compare to other products?

[1016] In this way, users can interactively obtain product information while visually checking the interactions between AI avatars.

[1017] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1018] Step 1:

[1019] The server receives the user's feature data and uses TensorFlow to build a generative AI model. The input is the user's feature data, and the output is an AI avatar corresponding to the user. Specifically, the server analyzes the user's profile information and past behavioral data and generates the AI ​​avatar based on that.

[1020] Step 2:

[1021] The server matches the generated AI avatars with each other. The input is the generated multiple AI avatars, and the output is the matchmaking results data. Specifically, the AI ​​avatars converse with each other, and the content of the conversation is analyzed to evaluate compatibility.

[1022] Step 3:

[1023] The server determines the compatibility between users based on the matchmaking results. The input is matchmaking result data, and the output is compatibility evaluation data. Specifically, it analyzes emotions in the conversation and extracts commonalities to calculate a compatibility score.

[1024] Step 4:

[1025] The server generates scenarios in which AI avatars introduce and compare products. The input is the user's prompt, and the output is a product introduction or comparison scenario. Specifically, when a user inputs "Tell me about the features of my new smartphone," the AI ​​avatar generates a scenario that explains the product's features based on the question.

[1026] Step 5:

[1027] The device uses Unity to display animated expressions and movements of an AI avatar. The input is a product introduction or comparison scenario, and the output is an animated display. Specifically, the device renders the expressions and movements of the AI ​​avatar in real time as it explains the product's features.

[1028] Step 6:

[1029] The device uses OpenCV to track the user's gaze and facial expressions in real time, providing an interactive experience. The input is the user's gaze and facial expression data, and the output is interactive feedback. Specifically, an AI avatar provides detailed explanations of products that the user shows interest in.

[1030] Step 7:

[1031] Users can ask questions to the AI ​​avatar. The input is the user's question, and the output is the AI ​​avatar's answer. Specifically, when a user asks, "Please tell me a review of this product," the server generates an answer to the question and sends it to the device.

[1032] Step 8:

[1033] The server generates an answer to the question in real time and sends it to the user's device. The input is the user's question, and the output is the answer data. Specifically, the server uses a generative AI model to generate the optimal answer to the question and provides it to the user.

[1034] 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.

[1035] "Example 1"

[1036] One embodiment of the present invention is a means for generating an AI avatar incorporating an emotion engine. This emotion engine recognizes the user's emotions and reflects them in the AI ​​avatar's reactions. Specifically, when the user expresses emotions such as joy or sadness to the system, the emotion engine recognizes this and the AI ​​avatar responds accordingly. For example, if the user expresses joy, the AI ​​avatar will respond by smiling.

[1037] "Example 2"

[1038] Also, a method for matching AI avatars using an emotion engine is one embodiment of the present invention. When AI avatars are matched, the emotion engine recognizes the emotions shown by each AI avatar and determines the outcome of the match based on that. For example, if one AI avatar shows joy and the other AI avatar also shows joy, it is determined that they are compatible.

[1039] "Example 3"

[1040] Furthermore, one embodiment of the present invention also provides a means for providing feedback to the user on the AI ​​avatar's reactions using an emotion engine. The emotional reactions of the AI ​​avatar are visually fed back to the user. Specifically, if the AI ​​avatar shows joy, that reaction is displayed as a smile on the screen and fed back to the user. This allows the user to directly see how their own emotions are reflected in the AI ​​avatar.

[1041] The processing flow of each embodiment will be described below.

[1042] "Example 1"

[1043] Step 1: The user expresses their feelings to the system, for example, expressing emotions such as happiness or sadness through text, voice, or facial expressions.

[1044] Step 2: The emotion engine recognizes the emotion expressed by the user. This emotion recognition is done using technologies such as text analysis, speech analysis, and facial expression analysis.

[1045] Step 3: The AI ​​avatar receives input from the emotion engine and generates a corresponding reaction. For example, if the user expresses joy, the AI ​​avatar will respond by smiling.

[1046] "Example 2"

[1047] Step 1: The AI ​​avatars begin meeting with each other. During this meeting, each AI avatar will engage in conversation that reflects the user's characteristics.

[1048] Step 2: The emotion engine recognizes the emotions expressed by each AI avatar by analyzing the text, voice, and facial expressions generated by the AI ​​avatar.

[1049] Step 3: The emotion engine determines the outcome of the matchmaking based on the emotions recognized. For example, if one AI avatar shows joy and the other AI avatar also shows joy, it is determined that they are a good match.

[1050] "Example 3"

[1051] Step 1: The AI ​​avatar generates an emotional response. This emotional response is generated based on the user's emotions and the results of the matchmaking between the AI ​​avatars.

[1052] Step 2: The AI ​​avatar's emotional response is visually displayed. Specifically, if the AI ​​avatar expresses joy, that response is displayed as a smile on the screen.

[1053] Step 3: The displayed emotional response of the AI ​​avatar is fed back to the user, allowing the user to directly see how their own emotions are reflected in the AI ​​avatar.

[1054] Example 1

[1055] Next, a description will be given of Example 1 of Form 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."

[1056] Conventional AI avatar systems have difficulty fully reflecting the characteristics and emotions of users, and lack the means to determine the compatibility between generated AI avatars. This makes it impossible to accurately determine the compatibility between users, reducing the usefulness of the system.

[1057] 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.

[1058] In this invention, the server includes means for providing an interface for users to input their own characteristics, means for preprocessing the input characteristic data, means for generating an AI avatar using the preprocessed data, means for providing an interface for inputting user emotion data, means for analyzing the input emotion data and reflecting it in the AI ​​avatar, means for matching the generated AI avatars with each other, and means for determining compatibility between users based on the results of the matchmaking. This makes it possible to generate AI avatars that reflect the users' characteristics and emotions, and to accurately determine compatibility between users through matching between AI avatars.

[1059] "User" refers to an individual who uses the system to input their own characteristics and emotions to generate an AI avatar.

[1060] "Interface" refers to the part of a system that provides a screen or input means for users to input their characteristics and emotions.

[1061] "Characteristic data" refers to individual pieces of information entered by users, such as hobbies, values, and lifestyle.

[1062] "Preprocessing" refers to the process of converting input feature data into a format suitable for learning by an AI avatar.

[1063] An "AI avatar" is a virtual character that is generated to reflect the user's characteristics and emotions.

[1064] "Emotional data" refers to information about emotions such as joy, sadness, and anger entered by the user.

[1065] An "emotion engine" refers to software that analyzes input emotional data and reflects it in the reactions of an AI avatar.

[1066] "Matchmaking" refers to the process of having generated AI avatars interact with each other to determine their compatibility.

[1067] "Compatibility assessment" refers to the process of evaluating the compatibility between users based on the results of matchmaking.

[1068] This invention relates to a system that allows users to input their own characteristics and emotions, generates AI avatars based on the input, and matches the AI ​​avatars with each other. Specific embodiments of this system are described below.

[1069] System configuration

[1070] This system consists of three main elements: a server, a terminal, and a user.

[1071] server

[1072] The server has the following functions:

[1073] 1. Interface provision function: Provides an interface for users to input their own characteristics.

[1074] 2. Data preprocessing function: Preprocesses the input feature data and converts it into a format suitable for learning by the AI ​​avatar.

[1075] 3. AI avatar generation function: Generates an AI avatar using preprocessed data.

[1076] 4. Emotional data analysis function: Analyzes the user's emotional data and reflects it in the AI ​​avatar.

[1077] 5. Matchmaking function: Matchmaking between generated AI avatars.

[1078] 6. Compatibility assessment function: Determines the compatibility between users based on the results of matchmaking.

[1079] Terminal

[1080] A terminal is a device through which a user inputs feature data and emotion data. Specifically, this includes a PC, smartphone, tablet, etc.

[1081] User

[1082] Users are individuals who use the system to input their own characteristics and emotions to generate an AI avatar.

[1083] Program processing

[1084] The server receives and preprocesses the feature data entered by the user through the device. The preprocessed data is then used to generate an AI avatar using a generative AI model. The generated AI avatar reflects the user's features and, by incorporating an emotion engine, reacts according to the user's emotional data.

[1085] Specifically, the server does the following:

[1086] 1. Data collection: The characteristic data entered by the user through the terminal is sent to the server.

[1087] 2. Data processing: The server preprocesses the collected feature data and converts it into a format suitable for learning the AI ​​avatar.

[1088] 3. AI avatar generation: The server generates an AI avatar using the preprocessed data, incorporating an emotion engine.

[1089] 4. Emotion recognition and response: The server uses an emotion engine to recognize the user's emotions and reflect them in the AI ​​avatar.

[1090] Specific examples

[1091] Here are some examples of specific prompts:

[1092] Example prompt sentence:

[1093] Enter your user characteristics:

[1094] Hobbies: Reading

[1095] Values: Cherishing family

[1096] Lifestyle: Health-conscious

[1097] Emotion input example:

[1098] Tell us how you're feeling today:

[1099] joy

[1100] Using this prompt, the server can collect the user's characteristics and generate the AI ​​avatar's response through the emotion engine.

[1101] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1102] Step 1:

[1103] The user inputs the feature data

[1104] Users input their own characteristic data through the device's interface, which is designed to allow users to input detailed information such as hobbies, values, and lifestyle.

[1105] Input: User characteristic data (e.g., "Hobbies: Reading, Values: Cherishing family, Lifestyle: Health-conscious")

[1106] Output: Feature data input to the device

[1107] Step 2:

[1108] The device sends the input data to the server

[1109] The device sends the characteristic data entered by the user to the server. This data is sent using a secure communication protocol (e.g., HTTPS).

[1110] Input: Feature data entered into the device

[1111] Output: Feature data sent to the server

[1112] Step 3:

[1113] The server preprocesses the data

[1114] The server preprocesses the received feature data, specifically converting text data into numerical data and formatting it in a way that can be applied to machine learning models.

[1115] Input: Feature data sent to the server

[1116] Data processing: Convert text data into numerical data (e.g., "reading" into "hobby category: 1," "cherishing family" into "value category: 2," and "health-conscious" into "lifestyle category: 3")

[1117] Output: Preprocessed data

[1118] Step 4:

[1119] The server generates an AI avatar

[1120] The server uses the preprocessed data to generate an AI avatar, using a generative AI model to create an avatar that reflects the user's characteristics.

[1121] Input: Preprocessed data

[1122] Data calculation: Input quantified data into the generative AI model to generate an AI avatar

[1123] Output: Generated AI avatar

[1124] Step 5:

[1125] The user inputs emotion data

[1126] A user inputs his / her own emotional data through the terminal interface. The emotional data indicates information about emotions such as joy, sadness, and anger.

[1127] Input: User's emotional data (e.g., "How are you feeling today?: Happy")

[1128] Output: Emotion data input to the device

[1129] Step 6:

[1130] The device sends emotion data to the server.

[1131] The device transmits the emotion data entered by the user to the server, which also uses a secure communication protocol.

[1132] Input: Emotion data entered into the device

[1133] Output: Emotion data sent to the server

[1134] Step 7:

[1135] The server analyzes the emotional data and reflects it in the AI ​​avatar.

[1136] The server analyzes the received emotional data and uses an emotion engine to reflect it in the AI ​​avatar, specifically changing the AI ​​avatar's facial expressions and reactions based on the emotional data.

[1137] Input: Emotion data sent to the server

[1138] Data calculation: Analyzes emotional data using the emotion engine and sets the AI ​​avatar's reaction (e.g., analyzes the emotional data of "joy" and sets the AI ​​avatar to smile)

[1139] Output: AI avatar that responds according to emotions

[1140] (Application example 1)

[1141] Next, a description will be given of Application Example 1 of Form 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."

[1142] While conventional systems could generate AI avatars that reflected the user's characteristics, they were unable to recognize the user's emotions and respond accordingly. Furthermore, they lacked the ability to recommend products based on the user's characteristics, limiting the shopping experience in virtual stores. This made it difficult to increase user satisfaction.

[1143] 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.

[1144] In this invention, the server includes means for generating an AI avatar that has learned the characteristics of each user, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for generating an AI avatar incorporating an emotion engine that recognizes the user's emotions and responds accordingly, means for recommending products based on the user's characteristics, and means for recognizing the user's emotions from camera footage.This makes it possible to respond according to the user's emotions and further recommend products based on the user's characteristics, thereby improving the shopping experience in virtual stores.

[1145] "User characteristics" refers to information specific to each individual user, such as the user's hobbies, values, and lifestyle.

[1146] An "AI avatar" is a virtual person or character created using artificial intelligence technology that reflects the user's characteristics.

[1147] An "emotion engine" is an artificial intelligence technology that recognizes the user's emotions and responds accordingly.

[1148] "Product recommendation" is the process of suggesting appropriate products based on the user's characteristics.

[1149] "Camera video" refers to video data acquired through a camera device.

[1150] "Matchmaking" is a process in which the generated AI avatars meet each other to determine their compatibility.

[1151] "Compatibility assessment" is the process of evaluating the compatibility between users based on the results of matchmaking.

[1152] The system for carrying out this invention consists of three main elements: a server, a terminal, and a user. The server includes means for generating an AI avatar that has learned the characteristics of each user, means for matching the generated AI avatars with each other, means for determining the compatibility between users based on the results of the matchmaking, means for generating an AI avatar incorporating an emotion engine that recognizes the user's emotions and responds accordingly, means for recommending products based on the user's characteristics, and means for recognizing the user's emotions from camera footage.

[1153] The terminal is a device such as a smartphone or head-mounted display, through which the user accesses the system. The terminal captures the user's image using a camera and sends it to the server. The server uses the captured image to perform emotion recognition and generates the reaction of the AI ​​avatar based on the results.

[1154] Specifically, the server processes and calculates data using software such as Python, OpenCV, Keras, and Scikit-learn. For example, it detects faces from camera footage and recognizes the user's emotions using an emotion recognition model. Depending on the recognized emotion, the AI ​​avatar responds by smiling or offering comforting words. It also converts the user's feature data into a TF-IDF vector and calculates the cosine similarity with a product database to recommend products.

[1155] As a concrete example, a user can access a virtual store using their smartphone, and an AI avatar can recommend an eco-bag to them. When the user smiles at the camera, the AI ​​avatar can respond by saying, "What a lovely smile!"

[1156] Examples of prompts for a generative AI model might include:

[1157] Input user characteristic data (hobbies, values, lifestyle, etc.). Next, input product database. Generate an AI avatar that recommends products based on the user's characteristics, recognizes the user's emotions from camera footage, and responds accordingly.

[1158] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1159] Step 1:

[1160] A user accesses the system through a terminal.

[1161] Input: User characteristic data (hobbies, values, lifestyle, etc.)

[1162] Output: User feature data is sent to the server

[1163] Specific operation: The user uses a smartphone or head-mounted display to access the system's interface and input their own characteristic data.

[1164] Step 2:

[1165] The server learns the user's characteristic data and generates an AI avatar.

[1166] Input: User feature data

[1167] Output: AI avatar that reflects the user's characteristics

[1168] Specific operation: The server uses Python to analyze the user's feature data and generates an AI avatar using a generative AI model.

[1169] Step 3:

[1170] The server acquires camera footage and performs emotion recognition.

[1171] Input: Camera image

[1172] Output: User emotion data

[1173] Specific operation: The device camera captures the user's image, detects the face using OpenCV, and recognizes the user's emotions using the Keras emotion recognition model.

[1174] Step 4:

[1175] The server uses an emotion engine to generate responses for the AI ​​avatar.

[1176] Input: User emotion data

[1177] Output: AI avatar response

[1178] Specific operation: Based on the recognized emotional data, the server generates a reaction from the AI ​​avatar, such as a smile or words of comfort.

[1179] Step 5:

[1180] The server recommends products using the user's feature data.

[1181] Input: User feature data, product database

[1182] Output: Recommended products list

[1183] Specific operation: The server uses Scikit-learn to convert the user's feature data into a TF-IDF vector, calculates the cosine similarity with the product database, and recommends products.

[1184] Step 6:

[1185] The device displays the AI ​​avatar's response and recommended products to the user.

[1186] Input: AI avatar response, recommended product list

[1187] Output: AI avatar responses and product list displayed to the user

[1188] Specific operation: The terminal displays the AI ​​avatar's reaction and the recommended product list sent from the server to the user, allowing the user to check and enjoy the shopping experience.

[1189] Example 2

[1190] Next, a description will be given of Example 2 of Form 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."

[1191] Conventional systems have difficulty accurately assessing the compatibility between users, and do not take into account the content or emotions of the conversation. Therefore, there is a need for a system that can more accurately assess the compatibility between users and efficiently advance the conversation.

[1192] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1193] In this invention, the server includes means for generating information processing devices that have learned the characteristics of individual users, means for having the generated information processing devices interact with each other, means for determining the compatibility between users based on the results of the interaction, means for analyzing the content of the interaction, the selection of topics, and the speed of reactions, and means for recognizing emotions shown by the information processing devices using an emotion recognition engine and evaluating the compatibility. This makes it possible to more accurately evaluate the compatibility between users and efficiently advance the interaction.

[1194] An "information processing device" is a device equipped with artificial intelligence to learn the characteristics of a user and engage in dialogue.

[1195] "Dialogue" is a process in which information processing devices converse with each other and exchange information with each other.

[1196] "Compatibility" is an index that indicates the compatibility between users, derived from the results of a dialogue between information processing devices.

[1197] The "contents of the conversation" refers to specific topics and words exchanged between information processing devices.

[1198] "Topic selection" refers to the theme or topic of conversation selected by the information processing device in a dialogue.

[1199] The "speed of response" refers to the time it takes for an information processing device to respond to a statement made by the other party in a conversation.

[1200] An "emotion recognition engine" is software or hardware for analyzing and recognizing emotions displayed by an information processing device.

[1201] "Emotion" refers to a psychological state such as joy, sadness, surprise, etc. that an information processing device shows during a conversation.

[1202] This invention is a system in which a server generates information processing devices (AI avatars) using a generative AI model and allows the generated information processing devices to interact with each other. A specific embodiment of this system is described below.

[1203] The server first generates an information processing device using a generative AI model (e.g., a natural language processing model). For example, a general natural language processing model is used as the generative AI model. The generated information processing device is set with individual personalities and topical tendencies.

[1204] Next, the server allows the generated information processing devices to interact with each other. The interaction takes the form of a conversation between the information processing devices. The server analyzes the content of the interaction, the choice of topics, and the speed of reactions in real time. This is done using natural language processing technology and machine learning algorithms.

[1205] Furthermore, the server recognizes emotions displayed by the information processing device using an emotion recognition engine. For example, general emotion analysis software is used for this emotion recognition engine. The server analyzes the emotions displayed by the information processing device using the emotion recognition engine and evaluates compatibility based on the analyzed emotions.

[1206] As a concrete example, the following prompt sentence is input to the generative AI model:

[1207] Example prompt sentence:

[1208] AI Avatar A: Hello, what would you like to talk about today?

[1209] AI Avatar B: Hello! I'd like to talk about a book I read recently.

[1210] AI Avatar A: That's great. What books have you read?

[1211] AI Avatar B: I recently read a book about the future of AI, and it was very interesting.

[1212] The server analyzes the flow of the conversation, the choice of topics, and the speed of reactions to calculate compatibility. It also uses an emotion recognition engine to recognize the emotions displayed by information processing devices A and B and reflect this in the compatibility evaluation. For example, if both devices show joy, the server determines that this is a good compatibility.

[1213] In this way, the server comprehensively evaluates the results of the dialogue between information processing devices, and the system more accurately judges the compatibility between users. This system enables users to efficiently find compatible partners.

[1214] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1215] Step 1:

[1216] The server generates an information processing device (AI avatar) using the generative AI model.

[1217] Input: A prompt for the generation AI model (e.g., "Generate a new AI avatar. It should have a cheerful personality and be interested in technology.")

[1218] Data processing: The generative AI model generates an AI avatar based on the prompt text and sets individual personality and topical tendencies.

[1219] Output: Generated AI avatar (information processing device)

[1220] Specific behavior:

[1221] The server inputs a prompt sentence into the generative AI model and generates an AI avatar, which is set with the specified personality and topical tendencies.

[1222] Step 2:

[1223] The server allows the generated information processing devices to interact with each other.

[1224] Input: Generated AI avatar (information processing device)

[1225] Data processing: The server initiates a dialogue between the AI ​​avatars and records the content of the conversation.

[1226] Output: Conversation log between AI avatars

[1227] Specific behavior:

[1228] The server pairs AI avatar A with AI avatar B and sends them an initial prompt. For example, the following conversation might take place:

[1229] AI Avatar A: Hello, what would you like to talk about today?

[1230] AI Avatar B: Hello! I'd like to talk about a book I read recently.

[1231] The server records this interaction.

[1232] Step 3:

[1233] The server analyzes the content of the conversation, the choice of topics, and the speed of responses.

[1234] Input: Conversation log between AI avatars

[1235] Data processing: Analyze the content of the conversation using natural language processing technology to extract topic selection and reaction speed.

[1236] Output: Analysis results (conversation topics, reaction times, etc.)

[1237] Specific behavior:

[1238] The server analyzes the interaction log and extracts data such as:

[1239] Conversation topics (e.g., technology, books)

[1240] Reaction time (e.g. AI avatar A's reaction time: 2 seconds, AI avatar B's reaction time: 3 seconds)

[1241] Step 4:

[1242] The server calculates the compatibility between AI avatars based on the analyzed data.

[1243] Input: Analysis results (conversation topics, reaction times, etc.)

[1244] Data processing: Compatibility is evaluated using statistical methods and machine learning algorithms.

[1245] Output: Compatibility score

[1246] Specific behavior:

[1247] The server calculates compatibility using the following algorithm:

[1248] Compatibility score = (match in conversation flow + match in topic selection + match in reaction speed) / 3

[1249] Step 5:

[1250] The server uses an emotion recognition engine to recognize emotions shown by the information processing device and evaluates compatibility.

[1251] Input: Conversation log between AI avatars

[1252] Data processing: Emotional data is extracted using an emotion recognition engine and reflected in compatibility evaluation.

[1253] Output: Emotion data, final compatibility evaluation

[1254] Specific behavior:

[1255] The server uses an emotion recognition engine to extract emotion data such as:

[1256] AI Avatar A's Emotion: Joy (80%)

[1257] AI Avatar B's Emotion: Joy (75%)

[1258] The server reflects this emotional data in the compatibility evaluation.

[1259] Step 6:

[1260] The server integrates the compatibility calculation results with the evaluation results of the emotion recognition engine and displays the final matchmaking results.

[1261] Input: compatibility score, emotional data

[1262] Data processing: Compatibility scores and emotional data are integrated to make the final evaluation.

[1263] Output: Final matchmaking results

[1264] Specific behavior:

[1265] The server displays the result to the user as follows:

[1266] The compatibility score between AI Avatar A and AI Avatar B is 85. Both are expressing happiness, which indicates good compatibility.

[1267] (Application example 2)

[1268] Next, a description will be given of Application Example 2 of Form 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."

[1269] Conventional matchmaking systems using AI avatars have had the problem of being difficult to accurately assess compatibility between users. Furthermore, because compatibility is assessed without taking emotional reactions into account, it often fails to reflect compatibility in real-life communication. Furthermore, there was a lack of means to assess compatibility through conversation in a virtual space, making it difficult to provide users with a realistic experience.

[1270] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1271] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for recognizing the emotions of the AI ​​avatars using an emotion engine and evaluating the results of the matchmaking based on those emotions, means for the AI ​​avatars to converse with each other in a virtual space, and means for calculating compatibility by analyzing the content and progress of the conversation and emotional responses. This makes it possible to more accurately determine compatibility between users and provide a realistic communication experience.

[1272] An "AI avatar that has learned the characteristics of each individual user" is a virtual character created by artificial intelligence that learns the individual characteristics and preferences of each user and is based on that.

[1273] "Means for matching generated AI avatars" refers to a method for allowing generated AI avatars to interact with each other through conversation and determining compatibility based on the results.

[1274] "Method of determining compatibility between users based on matchmaking results" refers to a method of analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users based on that data.

[1275] An "emotion engine" is a technology that recognizes the emotions of an AI avatar and reacts and evaluates based on those emotions.

[1276] A "virtual space" is a computer-generated virtual environment in which users can engage in activities and interactions.

[1277] "Method of calculating compatibility by analyzing the content, progress, and emotional responses of conversations" refers to a method of analyzing in detail the content, progress, and emotional responses of conversations between AI avatars, and calculating compatibility based on this data.

[1278] The system for carrying out this invention consists of three main elements: a server, a terminal, and a user. The server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for recognizing the emotions of the AI ​​avatars using an emotion engine and evaluating the results of the matchmaking based on those emotions, means for the AI ​​avatars to converse with each other in a virtual space, and means for calculating compatibility by analyzing the content and progress of the conversation and emotional responses.

[1279] The server uses OpenAI's API to generate conversations between AI avatars. It uses EmotionEngine to recognize emotions in the conversation and evaluates compatibility based on the results. Specifically, the server generates AI avatars based on data provided by users, and the avatars converse with each other in a virtual space. The content and progress of the conversation, as well as emotional responses, are analyzed to calculate compatibility.

[1280] The terminal is a device such as a smartphone or head-mounted display, and provides the interface for users to use the system. Through the terminal, users can create their own AI avatar and conduct matchmaking with other users' AI avatars. The progress of the matchmaking can be visually confirmed on the terminal screen.

[1281] Users input their characteristics and preferences into the system and use an AI avatar generated based on that information. The results of the matchmaking are fed back to the user, allowing them to enjoy communicating in a virtual space with compatible partners before actually meeting them in person.

[1282] As a concrete example, consider a scenario in which User A and User B go on a blind date in a virtual store. User A's AI avatar asks, "Hello, how was your day today?", and User B's AI avatar replies, "It was a great day. How about you?" The emotions in this conversation are analyzed and a compatibility score is calculated.

[1283] An example of a prompt is as follows:

[1284] User A's AI avatar: Hello, how was your day?

[1285] User B's AI avatar: It was a great day. How about you?

[1286] User A's AI avatar: I had a good day too. What are you interested in these days?

[1287] User B's AI avatar: I've been interested in reading lately. What about you?

[1288] In this way, it becomes possible to more accurately judge compatibility between users and provide a realistic communication experience.

[1289] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1290] Step 1:

[1291] The server receives feature data provided by the user and generates an AI avatar that learns the features of each individual user. The input is the user's feature data, and the output is the generated AI avatar. The server uses a generative AI model based on this data to create the AI ​​avatar.

[1292] Step 2:

[1293] The server matches the generated AI avatars with each other. The input is the generated multiple AI avatars, and the output is the matchmaking results. The server sets up the AI ​​avatars to converse with each other in a virtual space and records the content of the conversation.

[1294] Step 3:

[1295] The server uses an emotion engine to recognize the emotions of the AI ​​avatar. The input is the conversation data during the matchmaking session, and the output is the recognized emotion data. The server uses the Emotion Engine to analyze the emotions for each message in the conversation.

[1296] Step 4:

[1297] The server analyzes the content and progress of the conversation, as well as emotional responses, to calculate compatibility. The input is conversation data and emotional data, and the output is a compatibility score. The server calculates compatibility based on this data and saves the results.

[1298] Step 5:

[1299] The terminal visually displays the matchmaking results to the user. The input is the compatibility score and detailed matchmaking data, and the output is the results screen displayed to the user. The terminal receives this data and displays it in a format that is easy for the user to understand.

[1300] Step 6:

[1301] Users can check the results of their matchmaking through their device and enjoy communicating with compatible partners in a virtual space. The input is the results screen displayed on the device, and the output is the user's next action. The user selects the next step based on the results.

[1302] In this way, the server, terminal, and user can work together to more accurately determine compatibility between users and provide a realistic communication experience.

[1303] Example 3

[1304] Next, a description will be given of Example 3 of Form Example 3. 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."

[1305] In conventional AI avatar systems, there are limited means for determining compatibility between users, and a lack of visually verifiable information makes it difficult for users to fully evaluate the compatibility between avatars and the process. Furthermore, there is a lack of means for visually providing feedback on the avatar's emotional reactions, making it difficult for users to see how their own emotions are reflected in the avatar.

[1306] The identification process by the identification processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means. In this invention, the server includes means for generating an AI avatar that has learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining the compatibility between users based on the results of the matchmaking, means for generating conversation content using a generative AI model, means for displaying the generated conversation content as text, means for generating emotional responses of the avatars using an emotion engine, and means for displaying the generated emotional responses as animation. This allows users to visually check the conversation content and emotional responses between the avatars, and allows them to evaluate the compatibility between the avatars and the process in more detail.

[1307] An "artificial intelligence avatar" is a virtual character that can learn the user's characteristics and engage in conversation and express emotions.

[1308] "Means for generating" refers to a method or device for creating an AI avatar, including the process of having it learn the characteristics of the user.

[1309] "Means for matching" refers to a method or device for allowing the generated AI avatars to converse with each other and communicate with each other.

[1310] "Means for determining compatibility" refers to a method or device for evaluating the compatibility between users based on the results of matchmaking between artificial intelligence avatars.

[1311] A "generative artificial intelligence model" refers to a machine learning model for natural language processing and dialogue generation, including, for example, a large-scale language model.

[1312] "Means for generating conversation content" refers to a method or device for generating conversation content between AI avatars using a generative AI model.

[1313] The "means for displaying as text" refers to a method or device for displaying the generated conversation content as character information on the user's terminal.

[1314] "Emotion engine" refers to software and algorithms for generating emotional responses of artificial intelligence avatars, including functions for emotion recognition and emotional expression.

[1315] "Means for generating an emotional response" refers to a method or apparatus for using an emotion engine to generate an emotional response of an artificial intelligence avatar.

[1316] The term "means for displaying as animation" refers to a method or device for displaying animation on a user's terminal to visually express the generated emotional reaction.

[1317] This invention relates to a matchmaking system for AI avatars, which allows users to visually check the content of conversations and emotional reactions between the avatars. Specific embodiments of this system are described below.

[1318] System configuration

[1319] This system consists of three main components: a server, a terminal, and a user. The server generates AI avatars, generates conversation content, and generates emotional responses. The terminal displays the generated conversation content and emotional responses. The user accesses the system and initiates matchmaking between avatars.

[1320] Hardware and software used

[1321] Server: A server with high-performance computing capabilities is required. Specifically, a server equipped with a GPU is preferable.

[1322] Generative AI model: A large-scale language model (such as GPT-4) is used as a generative AI model for performing natural language processing.

[1323] Emotion Engine: Uses emotion recognition software (e.g., Affectiva's emotion recognition software) to generate emotional responses for the avatar.

[1324] Device: The device used by the user can be anything that can run a web browser or dedicated app, including PCs, smartphones, and tablets.

[1325] Program processing

[1326] The server first generates an AI avatar that learns the user's characteristics. To do this, it collects the user's input data and generates the avatar using a machine learning algorithm. The generated avatar reflects the user's characteristics and is capable of conversation and emotional expression.

[1327] Next, the server uses the generative AI model to generate conversation content between the avatars. It sends prompt text to the generative AI model and converts the returned conversation content into text format. For example, the following prompt text can be used:

[1328] "Please generate a conversation for a matchmaking session between two AI avatars. The conversation should be for Avatar A and Avatar B, where they introduce themselves to each other for the first time."

[1329] The generated conversation content is sent from the server to the device, which displays it on the screen, allowing the user to check the conversation between the avatars as text.

[1330] Furthermore, the server uses an emotion engine to generate an emotional response for the avatar. The generated emotional response is sent to the device as an animation, which is then displayed on the screen. For example, if avatar A expresses joy, the reaction is displayed as an animated smile.

[1331] Specific examples

[1332] A user accesses the system and clicks the "Start Matchmaking" button. The server sends the following prompt to the generative AI model:

[1333] "Please generate a conversation for a matchmaking session between two AI avatars. The conversation should be for Avatar A and Avatar B, where they introduce themselves to each other for the first time."

[1334] The conversation content returned by the generative AI model is as follows:

[1335] Avatar A: "Hi, how was your day today?"

[1336] Avatar B: "Hi, I had a great day. How about you?"

[1337] The server sends the conversation content as text to the device, which then displays it on the screen. If Avatar A expresses joy using the emotion engine, the server generates a smiling animation to represent the avatar's reaction and sends it to the device. The device then displays this smiling animation on the screen.

[1338] Users can view the conversation content and animated smiles of the avatars displayed on the screen, and evaluate the compatibility between the AI ​​avatars and the process.

[1339] In this way, this system allows users to visually check the content of conversations and emotional reactions between avatars, and evaluate the compatibility between avatars and the process in more detail.The flow of the identification process in Example 3 will be explained using Figure 21.

[1340] Step 1:

[1341] The user accesses the system and clicks the "Start Matchmaking" button. The input is the user's operation, and the output is a request to start matchmaking sent to the server. Specifically, the user accesses the system through a web browser or a dedicated app and clicks the "Start Matchmaking" button on the screen.

[1342] Step 2:

[1343] The server generates an AI avatar that has learned the user's characteristics. The user's characteristic data is sent to the server as input, and the AI ​​avatar is generated as output. Specifically, the server uses a machine learning algorithm to learn the user's characteristics and generates an AI avatar based on that.

[1344] Step 3:

[1345] The server uses a generative AI model to generate conversation content between avatars. A prompt sentence is sent to the generative AI model as input, and the conversation content is generated as output. Specifically, the server sends the following prompt sentence to the generative AI model:

[1346] "Please generate a conversation for a matchmaking session between two AI avatars. The conversation should be for Avatar A and Avatar B, where they introduce themselves to each other for the first time."

[1347] Step 4:

[1348] The server sends the generated conversation content to the terminal as text. The generated conversation content is input, and text format data is sent to the terminal as output. In concrete terms, the server converts the generated conversation content into text format and sends it to the user's terminal.

[1349] Step 5:

[1350] The device displays the conversation content on the screen. The input is text data sent from the server, and the output is displayed on the screen. Specifically, the device displays the received text data on the screen. For example, avatar A's comments are displayed on the left side, and avatar B's comments are displayed on the right side.

[1351] Step 6:

[1352] The server uses an emotion engine to generate the avatar's emotional response. The generated conversation content is the input, and emotional response data is generated as the output. In concrete terms, the server uses the emotion engine to generate the avatar's emotional response based on the conversation content. For example, if avatar A shows joy, the corresponding emotion data is generated.

[1353] Step 7:

[1354] The server sends the generated emotional reactions as animations to the device. The input is the generated emotional reaction data, and the output is animation data sent to the device. Specifically, the server generates animations of the avatar's facial expressions and movements based on the emotional data, and sends them to the user's device.

[1355] Step 8:

[1356] The device displays the animation on the screen. The input is animation data sent from the server, and the output is displayed on the screen. In concrete terms, the device displays the received animation data on the screen. For example, an animation of avatar A smiling is played.

[1357] Step 9:

[1358] The user visually checks the conversation and emotional reactions between the AI ​​avatars. The input is the conversation content and animation displayed on the screen, and the output is the user's evaluation. Specifically, the user checks the conversation content and the avatars' emotional reactions displayed on the screen, and evaluates the compatibility between the AI ​​avatars and the process.

[1359] (Application example 3)

[1360] Next, a description will be given of Application Example 3 of Form Example 3. 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."

[1361] Conventional AI avatar systems have limited means for determining compatibility between users, and there is a lack of visually verifiable information. Even in customer service systems using AI avatars, improving customer satisfaction is an issue because the AI ​​avatars do not adequately introduce products to customers or answer their questions. Furthermore, there is no way for users to visually see how the AI ​​avatars compare and review products, making it difficult for users to evaluate the process.

[1362] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1363] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for the AI ​​avatars to introduce products to customers and answer their questions, means for visually checking how the AI ​​avatars compare and review products, and means for visually providing feedback on the emotional reactions of the AI ​​avatars.This allows users to visually check how the AI ​​avatars match with each other and compare and review products, and enables the server to efficiently introduce products to customers and answer their questions.

[1364] "User" refers to any individual who uses the system.

[1365] "AI avatar" refers to a virtual person or character generated using artificial intelligence technology.

[1366] "Matchmaking" refers to the process by which AI avatars converse and interact with each other.

[1367] "Compatibility" refers to the criteria for evaluating compatibility and suitability between users.

[1368] "Customer" refers to a person who wishes to purchase an item in a virtual store.

[1369] "Product introduction" refers to the act of an AI avatar explaining the features and benefits of a product to customers.

[1370] "Answering questions" refers to the act of an AI avatar providing appropriate information in response to a customer's question.

[1371] "Product comparison" refers to the process of comparing the features and benefits of multiple products.

[1372] A "review" refers to the act of rating or expressing opinions about a product.

[1373] "Visually identifiable" refers to a state in which a user can visually confirm information on a screen.

[1374] "Emotional responses" refer to emotions such as joy or sadness shown by an AI avatar.

[1375] "Feedback" refers to the process by which a system returns information to a user.

[1376] The system for implementing this invention consists of three main elements: a server, a terminal, and a user.

[1377] server

[1378] The server runs a program that has the following functions:

[1379] 1. AI avatar generation

[1380] The server generates an AI avatar that learns the characteristics of each individual user, using the user's profile information and past behavioral data. The generated AI avatar is a virtual character that reflects the user's characteristics.

[1381] 2. Matchmaking simulation

[1382] The server then matches the generated AI avatars with each other. In this process, the AI ​​avatars converse and interact with each other, and the server determines the compatibility between users based on the results.

[1383] 3. Visual Feedback

[1384] The server allows users to visually check the matchmaking process between the AI ​​avatars by displaying the content of the conversation as text and animating the avatars' facial expressions and movements.

[1385] 4. Emotion Engine

[1386] The server uses an emotion engine to generate the AI ​​avatar's emotional response and provides visual feedback. For example, if the AI ​​avatar expresses joy, the response is displayed as a smile on the screen.

[1387] 5. Product introduction and comparison

[1388] The server will have the function of allowing the AI ​​avatar to introduce products to customers and answer their questions, and will also allow users to visually see how the AI ​​avatars compare and review products with each other.

[1389] Terminal

[1390] A terminal is a device through which a user interacts with the system, including a smartphone, smart glasses, or a head-mounted display.

[1391] 1. Display function

[1392] The device displays visual feedback sent from the server, including the AI ​​avatar's speech, facial expressions, and movements.

[1393] 2. Input Function

[1394] The terminal accepts input from the user, including text input and voice input.

[1395] User

[1396] The user is an individual human who uses the system. The user interacts with an AI avatar through a terminal.

[1397] Specific examples

[1398] For example, if a user types "Tell me about this product," the server uses a generative AI model to generate an appropriate answer and provides feedback to the user through an AI avatar. Examples of prompt sentences include:

[1399] Generate an emotional response to the following text: Tell me about this product.

[1400] In this way, users can learn more about products through AI avatars, and can make better purchasing decisions by visually watching AI avatars compare and review products.

[1401] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1402] Step 1:

[1403] The server receives user profile information and past behavioral data as input and generates an AI avatar that learns the characteristics of each individual user. In this process, the server analyzes the data using a generative AI model and outputs a virtual character that reflects the user's characteristics.

[1404] Step 2:

[1405] The server matches the generated AI avatars with each other. In this step, it generates prompts for the AI ​​avatars to converse and interact with each other, and simulates a conversation based on these prompts. The input is the generated AI avatar and the prompt, and the output is a conversation log resulting from the matchmaking.

[1406] Step 3:

[1407] The server determines the compatibility between users based on the matchmaking results. It analyzes the conversation log as input and outputs a compatibility score. This analysis uses natural language processing technology to evaluate the emotions and tone of the conversation.

[1408] Step 4:

[1409] The server allows users to visually check the matchmaking process between the AI ​​avatars. Specifically, it displays the conversation content as text and animates the avatars' facial expressions and movements. The input is the conversation log, and the output is visual feedback.

[1410] Step 5:

[1411] The server uses an emotion engine to generate the AI ​​avatar's emotional response and provides visual feedback. For example, if the AI ​​avatar expresses joy, the response is displayed as a smile on the screen. The input is the conversation log and the prompt sentence of the emotion engine, and the output is an animation of the emotional response.

[1412] Step 6:

[1413] The server executes the function of an AI avatar to introduce products to customers and answer their questions. It receives questions from users as input, uses a generative AI model to generate appropriate answers, and provides feedback to the user through the AI ​​avatar. The output is detailed product information.

[1414] Step 7:

[1415] The server allows users to visually see how AI avatars compare and review products. It receives multiple product data as input, generates comparison results and reviews, and displays them visually. The output is the product comparison results and reviews.

[1416] Step 8:

[1417] The device displays visual feedback sent from the server, including the AI ​​avatar's conversation content, facial expressions, and movements. The input is the feedback data from the server, and the output is the visual display for the user.

[1418] Step 9:

[1419] The terminal accepts input from the user, which can be text or voice input. The input is the user's question or instruction, and the output is the request data to the server.

[1420] Step 10:

[1421] Users interact with the AI ​​avatar through their device, which allows them to obtain detailed product information, compare products with other AI avatars, and visually check reviews. The input is visual feedback from the device, and the output is the user's understanding and purchasing decision.

[1422] 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.

[1423] 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> ) 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.

[1424] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are listed.

[1425] 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.

[1426] [Third embodiment]

[1427] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1428] 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.

[1429] 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).

[1430] 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.

[1431] 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.

[1432] 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).

[1433] 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.

[1434] 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.

[1435] 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.

[1436] 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.

[1437] 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.

[1438] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1439] "Example 1"

[1440] In one embodiment of the present invention, a system is provided that includes an interface through which a user can input their own characteristics. This interface is designed to allow detailed input of individual characteristics, such as the user's hobbies, values, and lifestyle. The input characteristics are used as learning data for an AI avatar, and each AI avatar reflects the user's characteristics.

[1441] "Example 2"

[1442] Next, the generated AI avatars are matched with each other. This matchmaking takes place in the form of a conversation between the AI ​​avatars, and compatibility is judged based on the content and progress of the conversation. Specifically, the flow of the conversation between the AI ​​avatars, choice of topics, speed of reaction, etc. are analyzed, and compatibility is calculated from these elements.

[1443] "Example 3"

[1444] Furthermore, users can visually check the progress of the matchmaking between the AI ​​avatars. Specifically, the conversation between the AI ​​avatars is displayed as text, and the avatars' facial expressions and movements are animated. This allows users to evaluate not only the compatibility between the AI ​​avatars, but also the process itself.

[1445] The processing flow of each embodiment will be described below.

[1446] "Example 1"

[1447] Step 1: The user accesses the system and enters their characteristics. Characteristics are items in which users can enter detailed information about their individual characteristics, such as hobbies, values, and lifestyle.

[1448] Step 2: The system generates an AI avatar based on the input characteristics. This AI avatar will reflect the user's characteristics.

[1449] "Example 2"

[1450] Step 1: The generated AI avatars begin meeting with each other. The meeting takes place in the form of a conversation between the AI ​​avatars.

[1451] Step 2: The system analyzes the flow of conversation between the AI ​​avatars, their choice of topics, and the speed of their reactions, and calculates compatibility based on these factors.

[1452] "Example 3"

[1453] Step 1: The user visually checks the matchmaking process between the AI ​​avatars through the system.

[1454] Step 2: Specifically, the content of conversations between AI avatars is displayed as text, and the avatars' facial expressions and movements are displayed as animations.

[1455] Step 3: This allows users to evaluate not only the compatibility between AI avatars, but also the process.

[1456] Example 1

[1457] Next, a description will be given of Example 1 of Form 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."

[1458] Conventional AI avatar generation systems have had difficulty efficiently generating AI avatars that reflect the characteristics of users and determining compatibility between users. Furthermore, there was a lack of a way to visually check the matchmaking process between the generated AI avatars, making it less convenient for users. Furthermore, there were also issues with the low reusability of AI avatars and poor time efficiency.

[1459] The identification process by the identification processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes a means having an interface for inputting the characteristics of individual users, a means for saving the input characteristics in a database, a means for creating a training dataset based on the saved data, a means for generating an AI avatar using a generative AI model, a means for matching the generated AI avatars with each other, and a means for determining the compatibility between users based on the results of the matchmaking. This makes it possible to efficiently generate AI avatars that reflect the characteristics of users and accurately determine the compatibility between users. In addition, the matchmaking process between the generated AI avatars can be visually confirmed, improving convenience for users. Furthermore, the reusability of AI avatars is increased, and time efficiency is improved.

[1460] "User" refers to an individual who uses the system to input their own characteristics and generate an AI avatar.

[1461] "Interface" refers to the means by which a screen or form is presented to a user to input their characteristics.

[1462] "Characteristics" refers to individual attributes and information such as a user's hobbies, values, and lifestyle.

[1463] "Database" refers to an information system for storing and managing input characteristics.

[1464] A "training dataset" refers to a collection of data used to train an AI model based on stored feature data.

[1465] A "generative AI model" refers to an algorithm or program for generating an AI avatar based on a training dataset.

[1466] An "AI avatar" refers to a virtual character generated to reflect the user's characteristics.

[1467] "Matchmaking" refers to the process by which generated AI avatars interact with each other and determine their compatibility.

[1468] "Compatibility" refers to an indicator that shows how closely users' characteristics and values ​​match.

[1469] "Means for visual confirmation" refers to a function that allows users to view the matchmaking process between the generated AI avatars on the screen.

[1470] "Reusability" refers to the property that an AI avatar that has been trained can be used any number of times.

[1471] "Time efficiency" refers to the speed at which the AI ​​avatar generation and matchmaking process is carried out.

[1472] The present invention provides a system having an interface for allowing a user to input his / her own characteristics. A specific embodiment of this system will be described below.

[1473] First, the user enters their characteristics using an interface on a web browser. This interface is built using HTML, CSS, and JavaScript. The user can enter detailed information about their individual characteristics, such as hobbies, values, and lifestyle. For example, the user might enter "Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious."

[1474] The terminal then sends the data entered by the user to the server using an HTTP POST request. This communication uses Axios, a JavaScript HTTP client library. The terminal sends the entered data to the server in JSON format.

[1475] The server analyzes the received data and saves it in a database. A relational database such as MySQL or PostgreSQL is used for the database. The server analyzes the received data and saves it in the database as "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-Oriented."

[1476] Next, the server creates a training dataset based on the stored data. This dataset reflects the user's characteristics. The server creates a training dataset for the AI ​​model based on the data retrieved from the database.

[1477] The server uses the generated dataset to generate an AI avatar. This process uses a generative AI model (such as GPT-3 or BERT). The server uses the generative AI model to generate an AI avatar with the following characteristics: Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious.

[1478] Finally, the device displays the AI ​​avatar received from the server. A front-end framework such as React.js is used for display. The device displays the AI ​​avatar received from the server using React.js, showing the user, "Hello, I'm a health-conscious avatar who loves reading and values ​​family."

[1479] As a specific example, if a user enters "Hobbies: Reading, Values: Valuing family, Lifestyle: Health-conscious," the server will generate an AI avatar based on this data and display to the user, "Hello, I'm a health-conscious avatar who loves reading and values ​​family."

[1480] An example prompt is, "Based on the user's hobbies, values, and lifestyle, please generate an AI avatar that reflects their individual characteristics. For example, if their hobby is reading, their values ​​are family-oriented, and their lifestyle is health-conscious, please create an AI avatar with those characteristics."

[1481] In this way, the system can efficiently generate AI avatars that reflect the characteristics of each user and determine compatibility between users. Furthermore, users can visually check the matchmaking process between the generated AI avatars, improving convenience for users. Furthermore, the reusability of AI avatars is enhanced, improving time efficiency.

[1482] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1483] Step 1:

[1484] User enters characteristics

[1485] A user inputs their own characteristics using an interface on a web browser. This interface is built using HTML, CSS, and JavaScript. Specifically, the user inputs "Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" into the form and clicks the submit button. The input data includes individual characteristics such as hobbies, values, and lifestyle.

[1486] Step 2:

[1487] The device sends the data to the server

[1488] The device sends the data entered by the user to the server using an HTTP POST request. This communication uses Axios, a JavaScript HTTP client library. Specifically, the device sends the data "Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" in JSON format to the server. The input is the feature data entered by the user, and the output is the JSON data sent to the server.

[1489] Step 3:

[1490] The server receives and stores the data

[1491] The server analyzes the received data and stores it in a database. A relational database such as MySQL or PostgreSQL is used for the database. Specifically, the server analyzes the received data and stores it in the database as "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-Oriented." The input is JSON data sent from the device, and the output is the user feature data stored in the database.

[1492] Step 4:

[1493] The server creates the training dataset.

[1494] The server creates a training dataset based on the stored data. This dataset reflects the user's characteristics. Specifically, the server retrieves the following data from the database: "User ID: 123, Hobbies: Reading, Values: Cherishing Family, Lifestyle: Health-conscious" and creates a training dataset for the AI ​​model. The input is the user characteristic data retrieved from the database, and the output is the training dataset.

[1495] Step 5:

[1496] The server generates an AI avatar

[1497] The server uses the generated dataset to generate an AI avatar. This process uses a generative AI model (such as GPT-3 or BERT). Specifically, the server uses the generative AI model to generate an AI avatar with the following characteristics: Hobbies: Reading, Values: Caring for family, Lifestyle: Health-conscious. The input is the training dataset, and the output is the generated AI avatar.

[1498] Step 6:

[1499] The device displays an AI avatar

[1500] The device displays the AI ​​avatar received from the server. A front-end framework such as React.js is used for display. Specifically, the device displays the AI ​​avatar received from the server using React.js, showing the user "Hello, I'm a health-conscious avatar who loves reading and values ​​family." The input is the AI ​​avatar received from the server, and the output is the AI ​​avatar displayed to the user.

[1501] (Application example 1)

[1502] Next, a description will be given of Application Example 1 of Form 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."

[1503] Conventional content distribution services have the problem of not providing sufficient personalized content recommendations based on users' hobbies, values, and lifestyles. Furthermore, there is a lack of systems that allow users to input their own detailed characteristics and then recommend the most appropriate content based on that information, which can lead to a decline in user satisfaction.

[1504] 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.

[1505] In this invention, the server includes a means for generating an AI avatar that has learned the characteristics of each user, a means for recommending optimal content to the user based on the generated AI avatar, and a means for providing the recommended content to the user. This enables personalized content recommendations based on the user's characteristics, thereby improving user satisfaction.

[1506] "User" refers to any individual who uses the system.

[1507] "Characteristics" refers to individual attributes such as a user's hobbies, values, and lifestyle.

[1508] An "AI avatar" refers to an artificial intelligence model that learns the user's characteristics and is generated based on them.

[1509] "Content" refers to information and entertainment such as movies, music, and articles.

[1510] "Recommendation" refers to the act of selecting and presenting the most suitable content based on the user's characteristics.

[1511] "Providing" refers to the act of actually showing or making available the recommended content to a user.

[1512] "Visual confirmation means" refers to a method that allows users to visually confirm the recommendation process and results through a screen or other means.

[1513] "Time-efficient" refers to a state in which a goal can be achieved effectively in a short amount of time.

[1514] The system for implementing this invention comprises a server and a user terminal. The server includes a means for generating an AI avatar that has learned the characteristics of each user, a means for recommending optimal content to the user based on the generated AI avatar, and a means for providing the recommended content to the user.

[1515] Program processing explanation

[1516] 1. User characteristics input

[1517] The user terminal provides an interface for users to input their own characteristics (hobbies, values, lifestyle, etc.). This interface is implemented as a smartphone application. The characteristics input by the user are sent to the server.

[1518] 2. AI avatar generation

[1519] The server generates an AI avatar based on the received user feature data. This AI avatar is trained using a generative AI model. Specifically, it uses Python and the scikit-learn library to vectorize the user's features and save them as an AI avatar.

[1520] 3. Content Recommendations

[1521] The server then recommends the most suitable content to the user based on the generated AI avatar. The recommendation algorithm is implemented using TF-IDF vectorization and cosine similarity calculation, which selects the content that most closely matches the user's characteristics.

[1522] 4. Provision of Content

[1523] The server transmits the recommended content to the user terminal for visual confirmation by the user, and the user terminal provides an interface for displaying the recommended content.

[1524] Specific examples

[1525] For example, if a user enters the following characteristics:

[1526] Hobbies: Watching movies, listening to music

[1527] Values: Cherishing time with family

[1528] Lifestyle: Health-conscious

[1529] In this case, an example prompt to input to the generative AI model would be:

[1530] The user's hobbies are watching movies and listening to music, and they value spending time with their family. Their lifestyle is health-conscious. Please recommend the most suitable content for this user.

[1531] Based on this prompt, the server will recommend the most suitable content for the user, such as movies, music, articles, etc., and provide it to the user's device, allowing the user to enjoy personalized content based on their own characteristics.

[1532] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1533] Step 1:

[1534] The user device provides an interface where the user can input their own characteristics (hobbies, values, lifestyle, etc.). The characteristics input by the user are sent to the server in JSON format. Examples of input data include "Hobbies: watching movies, listening to music," "Values: valuing time with family," and "Lifestyle: health-conscious."

[1535] Step 2:

[1536] The server analyzes the received user feature data and generates an AI avatar using a generative AI model. Specifically, it uses Python and the scikit-learn library to vectorize the user features using TF-IDF and save them as an AI avatar. The input data are the user features, and the output data is the vectorized AI avatar.

[1537] Step 3:

[1538] The server recommends the most suitable content to the user based on the generated AI avatar. The recommendation algorithm is implemented using TF-IDF vectorization and cosine similarity calculation. Specifically, it compares the user's feature vector with the vector in the content list and selects the content with the highest similarity. The input data is the AI ​​avatar and the content list, and the output data is the recommended content.

[1539] Step 4:

[1540] The server transmits the recommended content to the user terminal. The user terminal provides an interface for displaying the recommended content. Specifically, the recommended content, such as movies, music, or articles, is displayed on the user's screen. The input data is the recommended content, and the output data is the content displayed on the user terminal.

[1541] Step 5:

[1542] The user can visually check and use the recommended content. When the user selects content, the selection information is fed back to the server and used for the next recommendation. The input data is the user's selection information, and the output data is the feedback information that will be reflected in the next recommendation.

[1543] Example 2

[1544] Next, a description will be given of Example 2 of Form 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."

[1545] Conventional matchmaking systems using AI avatars have had difficulty accurately determining compatibility between users. Furthermore, they lacked the functionality to analyze the content and progress of conversations in real time and calculate compatibility, making it difficult for users to obtain satisfactory results. Furthermore, there was no way to visually check the matchmaking process between the generated AI avatars, making it difficult for users to understand or accept the system.

[1546] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1547] In this invention, the server includes a means for a user to input a prompt sentence, a means for generating an AI avatar using a generative AI model, a means for matching the generated AI avatars, a means for analyzing the content and progress of the conversation, a means for calculating the compatibility between the AI ​​avatars, a means for notifying the user of the compatibility results, a means for the user to visually confirm the progress of the matchmaking between the AI ​​avatars, and a means for allowing the AI ​​avatar to be used repeatedly once trained, thereby allowing the matchmaking to proceed time-efficiently. This makes it possible to accurately determine the compatibility between users and provide satisfactory results. Furthermore, since users can visually confirm the progress of the matchmaking, understanding and satisfaction are improved.

[1548] A "prompt sentence" is an instruction sentence that a user inputs to a generative AI model to generate an AI avatar.

[1549] A "generative AI model" is an artificial intelligence model for generating an AI avatar based on an input prompt sentence.

[1550] An "AI avatar" is a virtual character generated by a generative AI model with specific characteristics and personality.

[1551] "Matchmaking" is a process in which generated AI avatars converse with each other to determine their compatibility.

[1552] "Means for analyzing the content and progress of conversations" refers to a function that monitors and evaluates in real time the flow of conversations between AI avatars, topic selection, speed of reactions, etc.

[1553] The "means for calculating compatibility" is a function that calculates the compatibility between AI avatars in the form of numbers, graphs, etc. based on the results of conversation analysis.

[1554] The "means for notifying compatibility results" is a function for notifying the user of the calculated compatibility results.

[1555] "Visual confirmation means" is a function that allows users to visually check the progress of matchmaking between AI avatars.

[1556] "A time-efficient means of conducting matchmaking" is a function that allows the AI ​​avatar to progress through matchmaking in a short amount of time, and can be used any number of times once it has been trained.

[1557] This invention is a system in which a user inputs a prompt sentence, generates an AI avatar using a generative AI model, and matches the generated AI avatars with each other. Specific embodiments of this system are described below.

[1558] First, the user inputs a prompt using a terminal. The prompt is an instruction for the generative AI model to generate an AI avatar, such as "You are a 30-year-old man whose hobbies are reading and hiking. Please introduce yourself."

[1559] The server then receives the prompt and generates an AI avatar using a generative AI model (e.g., OpenAI's GPT-4 or Google's BERT). The generated AI avatar has information such as a self-introduction and hobbies based on the prompt.

[1560] The server pairs the generated AI avatars and initiates matchmaking. Matchmaking takes place by having the AI ​​avatars converse with each other. The server monitors the content and progress of the conversation in real time and analyzes the following factors:

[1561] Conversation flow: Check that the conversation is flowing smoothly, i.e., that the topic is kept on topic and that the conversation is unfolding naturally.

[1562] Topic selection: Evaluate whether the topics chosen by the AI ​​avatar are appropriate, for example, ensuring that topics are based on shared interests.

[1563] Reaction speed: Measures how quickly the AI ​​avatar responds to what the other person says. Evaluates whether it responds quickly and appropriately.

[1564] The server analyzes these factors comprehensively and calculates the compatibility between the AI ​​avatars. The compatibility is displayed in the form of numbers and graphs. For example, if the compatibility is evaluated as 80%, the server notifies the user of the result.

[1565] Furthermore, the server provides a means for users to visually check the matchmaking process between the AI ​​avatars, allowing them to intuitively understand the progress and results of the matchmaking.

[1566] Furthermore, once the AI ​​avatar has been trained, it can be used any number of times, allowing for time-efficient matchmaking, enabling users to conduct multiple matchmaking sessions in a short amount of time and find the perfect partner.

[1567] The above is a specific embodiment for carrying out the present invention.

[1568] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1569] Step 1:

[1570] The user enters a prompt statement.

[1571] The user uses a terminal to input a prompt to the generative AI model to generate an AI avatar. For example, the prompt might be, "You are a 30-year-old man whose hobbies are reading and hiking. Please introduce yourself." The input prompt is then sent to the server.

[1572] Step 2:

[1573] The server generates an AI avatar using the generative AI model.

[1574] The server receives the prompt sent by the user and inputs it into a generative AI model (e.g., OpenAI's GPT-4). The generative AI model generates an AI avatar based on the prompt and determines the avatar's characteristics and personality. The data of the generated AI avatar is stored on the server.

[1575] Step 3:

[1576] The server matches the generated AI avatars with each other.

[1577] The server pairs the generated AI avatars and initiates matchmaking. Matchmaking takes place in the form of conversation between the AI ​​avatars. The server instructs each AI avatar to start a conversation and collects conversation logs in real time.

[1578] Step 4:

[1579] The server analyzes the content and progress of the conversation.

[1580] The server analyzes the collected conversation logs and evaluates the following factors:

[1581] Conversation flow: Check that the conversation is flowing smoothly, i.e., that the topic is kept on topic and that the conversation is unfolding naturally.

[1582] Topic selection: Evaluate whether the topics chosen by the AI ​​avatar are appropriate, for example, ensuring that topics are based on shared interests.

[1583] Reaction speed: Measures how quickly the AI ​​avatar responds to what the other person says. Evaluates whether it responds quickly and appropriately.

[1584] Step 5:

[1585] The server calculates the compatibility between AI avatars.

[1586] The server calculates the compatibility between the AI ​​avatars based on the data collected in step 4. The compatibility is displayed in the form of a number or a graph. For example, if the compatibility is evaluated as 80%, the server generates the result.

[1587] Step 6:

[1588] The server notifies the user of the compatibility results.

[1589] The server notifies the user of the calculated compatibility results, which the user can check via their device. This allows the user to decide their next action based on the results of the matchmaking.

[1590] Step 7:

[1591] To provide a means for a server to visually check the state of a matchmaking session.

[1592] The server provides an interface that allows users to visually check the matchmaking process between AI avatars, and users can intuitively understand the progress and results of the matchmaking process through their devices.

[1593] Step 8:

[1594] The server makes the AI ​​avatars reusable.

[1595] The server allows the generated AI avatar to be trained once and then used again and again, allowing users to conduct multiple matchmaking sessions in a short amount of time and find the perfect partner.

[1596] (Application example 2)

[1597] Next, a description will be given of Application Example 2 of Form 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."

[1598] Conventional matchmaking systems using AI avatars have had difficulty accurately determining compatibility between users. They also lack the ability to take into account detailed factors such as the flow of conversation, topic selection, and speed of response, resulting in insufficient calculations of compatibility. Furthermore, there was a lack of a way for users to visually check the progress of the matchmaking, which hindered the user experience.

[1599] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining the compatibility between users based on the results of the matchmaking, means for analyzing the flow of conversation, topic selection, and reaction speed, and means for calculating compatibility. This makes it possible to more accurately determine the compatibility between users and visually check the matchmaking process.

[1600] "Means for generating an AI avatar that has learned the characteristics of each individual user" refers to a method or device for learning the individual characteristics and preferences of each user and generating an AI avatar based on that.

[1601] "Means for matching generated AI avatars" refers to a method or device for matching generated AI avatars through conversation.

[1602] "Means for determining compatibility between users based on matchmaking results" refers to a method or device for analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users based on the results.

[1603] "Means for analyzing conversation flow, topic selection, and speed of response" refers to methods and devices for analyzing the flow of conversations between AI avatars, the topics selected, the speed of response, etc.

[1604] The "means for calculating compatibility" refers to a method or device for quantifying the compatibility between users based on the results of conversation analysis.

[1605] "Means for visual confirmation" refers to a method or device that allows users to visually confirm the matchmaking process between AI avatars.

[1606] "A means for efficiently conducting matchmaking in a time-efficient manner" refers to a method or device that allows an AI avatar to be trained once and then used any number of times, allowing matchmaking to proceed efficiently in a short amount of time.

[1607] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The following explains how each element functions.

[1608] First, the server has a means for generating an AI avatar that has learned the characteristics of each individual user. This means generates an AI avatar using a generative AI model based on the user's input data (e.g., profile information and past conversation history). The generated AI avatar reflects the user's characteristics and has individual preferences and personality.

[1609] Next, the server has a means for matching the generated AI avatars with each other. This means generates prompts for the generated AI avatars to converse with each other, and simulates the conversation using the generative AI model. For example, the following prompts are used:

[1610] Avatar 1: Hello, how are you doing these days?

[1611] Avatar 2: I'm fine. How about you?

[1612] Start a conversation.

[1613] The server has a means for analyzing the flow of conversation, topic selection, and speed of response. This means analyzes the generated conversation text and evaluates elements such as the flow of conversation, topic selection, and speed of response. This allows data to be obtained for quantifying the quality of the conversation and compatibility.

[1614] Furthermore, the server has a means for calculating compatibility, which quantifies the compatibility between users based on the results of conversation analysis. For example, if the conversation flows smoothly and the choice of topics reflects common interests, a high compatibility score is calculated.

[1615] The device has a means for users to visually check the matchmaking process. This allows users to visually check the conversation between the AI ​​avatars in real time. For example, they can use a smartphone or head-mounted display to visually experience the conversation in a virtual space.

[1616] Users can control the AI ​​avatar through their device and check the results of their matchmaking. Users can customize their own avatar and check compatibility by talking with other users' avatars.

[1617] This system allows users to more accurately judge compatibility between each other and visually check the matchmaking process. Specifically, it provides a realistic conversation experience through conversation simulation using a generative AI model and conversation generation using prompt sentences.

[1618] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1619] Step 1:

[1620] The server receives input data from the user. The input data includes the user's profile information and past conversation history. Based on this data, a generative AI model is used to generate an AI avatar that learns the user's characteristics. The generated AI avatar reflects the user's individual preferences and personality. The input is the user's data, and the output is the generated AI avatar.

[1621] Step 2:

[1622] The server generates a prompt to match the generated AI avatars. The prompt is text to start a conversation between the AI ​​avatars. For example, it takes the form "Avatar 1: Hello, how are you?\nAvatar 2: I'm fine. How about you?\nLet's start the conversation." The input is the generated AI avatar, and the output is the prompt.

[1623] Step 3:

[1624] The server uses a generative AI model to simulate a conversation based on the prompt sentence. The generative AI model receives the prompt sentence as input and generates a conversation between AI avatars. The generated conversation is output in text format. The input is the prompt sentence, and the output is the generated conversation text.

[1625] Step 4:

[1626] The server analyzes the generated conversation text. The analysis includes elements such as the flow of the conversation, topic selection, and speed of response. These elements are evaluated to obtain data to quantify the quality of the conversation and compatibility. The input is the generated conversation text, and the output is the analysis results.

[1627] Step 5:

[1628] The server calculates the compatibility between users based on the analysis results. Compatibility is high when the conversation flows smoothly and the topic choices reflect common interests. Compatibility is quantified and fed back to the user. The input is the analysis results, and the output is a compatibility score.

[1629] Step 6:

[1630] The terminal allows the user to visually check the progress of the matchmaking. The terminal visually displays the generated conversation text, allowing the user to check the conversation in real time. The input is the generated conversation text, and the output is the visual display.

[1631] Step 7:

[1632] Users control the AI ​​avatar through their device and check the matchmaking results. Users can customize their own avatar and check compatibility by conversing with other users' avatars. The input is a compatibility score, and the output is user feedback.

[1633] Example 3

[1634] Next, a third embodiment of the third embodiment will be described. 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."

[1635] In conventional systems using AI avatars, there were limited means for determining compatibility between users, and the system was unable to visually confirm it. It was also difficult to evaluate the process of dialogue between AI avatars, making it less convenient for users. Furthermore, generating AI avatars and simulating dialogue took time, making it difficult to proceed efficiently.

[1636] The specific processing by the specific processing unit 290 of the data processing device 12 in the third embodiment is realized by the following means.

[1637] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for having the generated AI avatars converse with each other, means for determining compatibility between users based on the results of the conversation, means for the user to input a prompt sentence, means for simulating a conversation between the generated AI avatars, means for generating animation data of the conversation content and the avatars, means for transmitting the generated data to a terminal, means for the terminal to display the conversation content and avatar animation, and means for the user to evaluate the compatibility between the AI ​​avatars and the progress of the conversation. This allows the user to visually check the state of the conversation between the AI ​​avatars and evaluate the process. Furthermore, once the AI ​​avatars have been trained, they can be used repeatedly, allowing for time-efficient conversations.

[1638] An "artificial intelligence avatar" is a virtual character that learns the user's characteristics and is generated to simulate conversations and actions.

[1639] A "prompt" is a text of instructions or questions entered by a user, which serves as the basis for generating an AI avatar and simulating a dialogue.

[1640] "Dialogue" refers to conversations and communication between artificial intelligence avatars, and is simulated based on user input.

[1641] "Simulation" is the process of virtually recreating the interactions and behaviors of artificially intelligent avatars, and is done using a generative AI model.

[1642] "Animation data" is data that visually expresses the facial expressions and movements of an artificial intelligence avatar, and is displayed to the user in real time.

[1643] A "terminal" is a device that allows a user to input prompts and visually confirm the dialogue and animation of an artificial intelligence avatar.

[1644] A "generative AI model" is a machine learning model used to generate artificial intelligence avatars and simulate dialogue, and includes, for example, natural language processing models.

[1645] "Compatibility" refers to the suitability and compatibility between users, as determined based on the results of conversations between AI avatars.

[1646] "Evaluation" refers to the judgment or feedback that the user gives based on the process and results of the dialogue between AI avatars.

[1647] The present invention is a system that uses an artificial intelligence avatar and provides a means for determining compatibility between users. A specific embodiment of this system will be described below.

[1648] First, the user inputs a prompt using the terminal. The prompt specifies the theme and content of the conversation between the AI ​​avatars. For example, the user can input a prompt such as "Start a matchmaking session between AI avatars discussing hobbies."

[1649] The server then receives the prompt and generates an AI avatar using a generative AI model, such as the natural language processing model GPT-4. The generated AI avatar is configured to have different personalities and hobbies based on the prompt.

[1650] The server simulates a conversation between the generated AI avatars. The content of the simulated conversation is generated in text format. For example, Avatar A might say, "I like reading. I especially like mystery novels," and Avatar B might reply, "I like reading, too. I've been hooked on historical novels lately."

[1651] Furthermore, the server generates animation data for the avatars' facial expressions and movements based on the content of the conversation. For example, it generates an animation of avatar A smiling and talking, and avatar B nodding.

[1652] The generated text data and animation data are sent from the server to the device. Based on the received data, the device displays the dialogue content as text to the user and animates the avatars' facial expressions and movements. This allows the user to visually confirm the state of the dialogue between the AI ​​avatars.

[1653] Finally, the user can evaluate the compatibility and dialogue process between the AI ​​avatars based on the dialogue content and animation displayed on the device. For example, they can evaluate that "Avatar A and Avatar B seem to get along well."

[1654] This system allows users to visually check the state of conversations between AI avatars and evaluate the process. Furthermore, once the AI ​​avatars have been trained, they can be used any number of times, allowing for time-efficient conversations.

[1655] As an example, the following prompt sentence is used:

[1656] "Starting matchmaking with AI avatars that discuss hobbies"

[1657] "Starting a matchmaking session with AI avatars to discuss travel"

[1658] "We're starting a matchmaking session for AI avatars that discuss music."

[1659] The above is an embodiment of the present invention. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1660] Step 1:

[1661] The user enters a prompt statement.

[1662] The user enters a prompt into the input field of the device, for example, "Start a matchmaking session for an AI avatar to discuss hobbies." The entered prompt is sent from the device to the server.

[1663] Step 2:

[1664] The server receives the prompt sentence and generates an artificial intelligence avatar using a generative AI model.

[1665] The server receives a prompt sent by the user. It then uses a generative AI model (e.g., a natural language processing model) to generate two AI avatars based on the prompt. The generated avatars are configured to have different personalities and hobbies. The input is the prompt, and the output is the generated AI avatars.

[1666] Step 3:

[1667] The server simulates conversations between artificial intelligence avatars.

[1668] The server simulates a conversation between the generated AI avatars. Using the generative AI model, Avatar A and Avatar B have a conversation based on a prompt. For example, Avatar A says, "I like reading," and Avatar B replies, "I like reading, too." The input is the generated AI avatar, and the output is the simulated conversation content.

[1669] Step 4:

[1670] The server generates the conversation content and avatar animation data.

[1671] The server generates the simulated conversation content in text format. It also generates the avatars' facial expressions and movements as animation data. For example, it generates an animation of avatar A smiling and talking, and avatar B nodding. The input is the simulated conversation content, and the output is text data and animation data.

[1672] Step 5:

[1673] The server transmits the generated data to the terminal.

[1674] The server sends the generated text data and animation data of the conversation to the terminal. The input is the text data and animation data, and the output is sent to the terminal.

[1675] Step 6:

[1676] The device displays the conversation and an animated avatar.

[1677] The device displays the conversation content as text to the user based on the data received from the server. It also displays the avatar's facial expressions and movements as animation. For example, avatar A and avatar B are displayed on the screen, and the conversation content and facial expressions of each are displayed in real time. The input is the data received from the server, and the output is what is displayed to the user.

[1678] Step 7:

[1679] Users evaluate the compatibility between AI avatars and the conversation process.

[1680] The user checks the conversation content and animation displayed on the device and evaluates the compatibility between the AI ​​avatars and the progress of the conversation. For example, the user may evaluate that "Avatar A and Avatar B seem to get along well." The input is the displayed conversation content and animation, and the output is the user's evaluation.

[1681] (Application example 3)

[1682] Next, a description will be given of Application Example 3 of Form Example 3. 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."

[1683] Conventional AI avatar systems have limited means for determining compatibility between users, and users cannot visually check the interactions between AI avatars. Furthermore, when introducing or comparing products, they lack the interactive functionality to allow users to ask questions in real time and receive answers. This has led to problems such as reduced user satisfaction and convenience.

[1684] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 3 is realized by the following means.

[1685] In this invention, the server includes means for generating AI avatars that have learned the characteristics of individual users, means for matching the generated AI avatars with each other, means for determining compatibility between users based on the results of the matchmaking, means for the AI ​​avatars to introduce and compare products with each other, means for visually checking this situation, and means for the user to ask the AI ​​avatar questions and receive responses in real time, thereby enabling users to interactively obtain product information while visually checking the interactions between the AI ​​avatars.

[1686] An "AI avatar" is a virtual character that learns the characteristics of each individual user and is generated based on those characteristics.

[1687] "Matchmaking" is a process in which generated AI avatars converse and interact with each other, and their compatibility is determined based on the results.

[1688] The "means of determining compatibility" is a method for analyzing the results of matchmaking between AI avatars and evaluating the compatibility between users.

[1689] "Means for introducing and comparing products" is a function that allows an AI avatar to provide information about specific products and compare multiple products.

[1690] "Visual confirmation" refers to a method that allows users to visually observe the interactions and animations between AI avatars.

[1691] "Real-time reception" is a feature that allows users to ask questions to an AI avatar and receive answers instantly.

[1692] A system for implementing the present invention includes a server, a user terminal, and an AI avatar generation program. Specific embodiments are described below.

[1693] System Configuration

[1694] 1. Server:

[1695] Hardware: A server equipped with a high-performance CPU, GPU, and memory.

[1696] Software: TensorFlow (generative AI model), Unity (animation display), OpenCV (image processing).

[1697] 2. User Device:

[1698] Hardware: Smartphone, Head-Mounted Display (HMD).

[1699] Software: Dedicated application (Unity-based).

[1700] Program processing

[1701] 1. AI Avatar Generation:

[1702] The server uses TensorFlow to build a generative AI model and have it learn the user's characteristics.

[1703] When a user enters a prompt sentence, the generative AI model generates an AI avatar based on its features.

[1704] 2. Conducting matchmaking:

[1705] The server matches the generated AI avatars with each other.

[1706] Based on the results of the matchmaking, the server determines the compatibility between users.

[1707] 3. Product introduction and comparison:

[1708] The server generates scenarios in which AI avatars introduce and compare products.

[1709] The user's device uses Unity to display the AI ​​avatar's facial expressions and movements in animated form.

[1710] 4. Visual Verification:

[1711] The user device uses OpenCV to track the user's gaze and facial expressions in real time, providing an interactive experience.

[1712] Users can visually see the interactions between AI avatars.

[1713] 5. Real-time Questions and Answers:

[1714] Users can ask questions to the AI ​​avatar.

[1715] The server generates an answer to the question in real time and transmits it to the user terminal.

[1716] Specific examples

[1717] The user puts on a smartphone or HMD and accesses a virtual store. An AI avatar appears on the screen, and the user enters a prompt: "Tell me about the features of my new smartphone." The AI ​​avatar responds to the question by comparing the features of several smartphones. The conversation is displayed as text, and the avatar's facial expressions and movements are animated, allowing the user to visually confirm the conversation.

[1718] Prompt Sentence Examples

[1719] "What's the difference between the latest smartphones and tablets?"

[1720] "Write a review for this product"

[1721] How does this product compare to other products?

[1722] In this way, users can interactively obtain product information while visually checking the interactions between AI avatars.

[1723] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1724] Step 1:

[1725] The server receives the user's feature data and uses TensorFlow to build a generative AI model. The input is the user's feature data, and the output is an AI avatar corresponding to the user. Specifically, the server analyzes the user's profile information and past behavioral data and generates the AI ​​avatar based on that.

[1726] Step 2:

[1727] The server matches the generated AI avatars with each other. The input is the generated multiple AI avatars, and the output is the matchmaking results data. Specifically, the AI ​​avatars converse with each other, and the content of the conversation is analyzed to evaluate compatibility.

[1728] Step 3:

[1729] The server determines the compatibility between users based on the matchmaking results. The input is matchmaking result data, and the output is compatibility evaluation data. Specifically, it analyzes emotions in the conversation and extracts commonalities to calculate a compatibility score.

[1730] Step 4:

[1731] The server generates scenarios in which AI avatars introduce and compare products. The input is the user's prompt, and the output is a product introduction or comparison scenario. Specifically, when a user inputs "Tell me about the features of my new smartphone," the AI ​​avatar generates a scenario that explains the product's features based on the question.

[1732] Step 5:

[1733] The device uses Unity to display animated expressions and movements of an AI avatar. The input is a product introduction or comparison scenario, and the output is an animated display. Specifically, the device renders the expressions and movements of the AI ​​avatar in real time as it explains the product's features.

[1734] Step 6:

[1735] The device uses OpenCV to track the user's gaze and facial expressions in real time, providing an interactive experience. The input is the user's gaze and facial expression data, and the output is interactive feedback. Specifically, an AI avatar provides detailed explanations of products that the user shows interest in.

[1736] Step 7:

[1737] Users can ask questions to the AI ​​avatar. The input is the user's question, and the output is the AI ​​avatar's answer. Specifically, when a user asks, "Please tell me a review of this product," the server generates an answer to the question and sends it to the device.

[1738] Step 8:

[1739] The server generates an answer to the question in real time and sends it to the user's device. The input is the user's question, and the output is the answer data. Specifically, the server uses a generative AI model to generate the optimal answer to the question and provides it to the user.

[1740] 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.

[1741] "Example 1"

[1742] One embodiment of the present invention is a means for generating an AI avatar incorporating an emotion engine. This emotion engine recognizes the user's emotions and reflects them in the AI ​​avatar's reactions. Specifically, when the user expresses emotions such as joy or sadness to the system, the emotion engine recognizes this and the AI ​​avatar responds accordingly. For example, if the user expresses joy, the AI ​​avatar will respond by smiling.

[1743] "Example 2"

[1744] Also, a method for matching AI avatars using an emotion engine is one embodiment of the present invention. When AI avatars are matched, the emotion engine recognizes the emotions shown by each AI avatar and determines the outcome of the match based on that. For example, if one AI avatar shows joy and the other AI avatar also shows joy, it is determined that they are compatible.

[1745] "Example 3"

[1746] Furthermore, one embodiment of the present invention also provides a means for providing feedback to the user on the AI ​​avatar's reactions using an emotion engine. The emotional reactions of the AI ​​avatar are visually fed back to the user. Specifically, if the AI ​​avatar shows joy, that reaction is displayed as a smile on the screen and fed back to the user. This allows the user to directly see how their own emotions are reflected in the AI ​​avatar.

[1747] The processing flow of each embodiment will be described below.

[1748] "Example 1"

[1749] Step 1: The user expresses their feelings to the system, for example, expressing emotions such as happiness or sadness through text, voice, or facial expressions.

[1750] Step 2: The emotion engine recognizes the emotion expressed by the user. This emotion recognition is done using technologies such as text analysis, speech analysis, and facial expression analysis.

[1751] Step 3: The AI ​​avatar receives input from the emotion engine and generates a corresponding reaction. For example, if the user expresses joy, the AI ​​avatar will respond by smiling.

[1752] "Example 2"

[1753] Step 1: The AI ​​avatars begin meeting with each other. During this meeting, each AI avatar will engage in conversation that reflects the user's characteristics. ...

Claims

1. means for receiving a prompt from a user; A means for inputting the prompt sentence into a generating AI model to generate an AI avatar that has learned the characteristics of an individual user; A means for causing the AI ​​avatars to converse with each other; A means for analyzing the conversation log and calculating the compatibility between the AI ​​avatars; and a means for notifying the user of the result of the compatibility calculation. system.

2. the means for calculating the compatibility extracts data relating to at least one of the conversation topic and the reaction time from the conversation log, and calculates the compatibility based on the extracted data; The system of claim 1 .

3. means for providing an interface for inputting emotion data of the user; and a means for analyzing the emotion data and reflecting the emotion data in the AI ​​avatar. The system of claim 1 .

Images