System

Abstract

To provide a system.SOLUTION: A system includes creation means for creating a profile photo of a user on the basis of an action history of the user and profile information by inputting a specific prompt sentence to a generative AI model, means for recommending a matching partner of the user about feelings by using a feeling engine for recognizing the user's feelings of the user and the generative AI model, means for tracking a change in the user's feelings, means for re-recommending the matching partner of the user on the basis of the change in the feelings, and means for monitoring matching states of all users, and recommending matching partners to users with a matching number that is less than a matching number of specific users.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] Conventional dating apps have faced issues such as uneven matching among users and the lack of matching skills. Additionally, human subjectivity plays a large role in creating user profile pictures and selecting potential partners, making efficient matching difficult. [Means for solving the problem]

[0005] The system of the present disclosure includes a creation means for creating a profile picture for the user based on the user's behavioral history and profile information by inputting a specific prompt sentence into a generative AI model; an emotion engine for recognizing the user's emotions; a means for recommending matching partners for the user based on the emotions using the generative AI model; a means for tracking changes in the user's emotions; a means for re-recommending matching partners for the user based on the changes in emotions; and a means for monitoring the matching status of all users and recommending matching partners for users whose number of matches is fewer than the number of matches of a specific user. [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] The system of the present invention utilizes AI technology to create a profile picture for a user. Specifically, the AI ​​generates the profile picture based on information such as a self-introduction, hobbies, and preferences provided by the user. For example, if a user provides information that they "like the outdoors," the AI ​​will set the background of the user's profile picture to an outdoor scene, such as a campsite or mountain.

[0029] "Example 2"

[0030] The system of the present invention also uses AI to perform advanced attribute judgments and guide potential matches. Specifically, the AI ​​analyzes the user's behavioral history and profile information to recommend optimal matches. For example, if a user tends to frequently communicate with people who share specific hobbies or preferences, the AI ​​can learn this tendency and recommend people with similar hobbies and preferences as potential matches. The profile information may include the user's attributes and the user's past matching history.

[0031] "Example 3"

[0032] Furthermore, the system of the present invention reduces the bias in the amount of matches. Specifically, the AI ​​monitors the matching status of all users and more proactively recommends potential matches to users with few matches. For example, if a user has not had a match for a certain period of time, the AI ​​will proactively recommend users who are likely to match the user's attributes.

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

[0034] "Example 1"

[0035] Step 1: The user provides information such as a self-introduction, hobbies, and preferences.

[0036] Step 2: The AI ​​generates a profile picture based on the provided information. For example, if a user provides information that they "like the outdoors," the AI ​​will set the user's profile picture with an outdoor scene, such as a campsite or mountain, as the background.

[0037] Step 3: The generated profile picture is set as the user's profile.

[0038] Step 1: The user's behavioral history and profile information are analyzed by AI. The profile information may include the user's attributes and past matching history.

[0039] Step 2: The AI ​​recommends optimal matches based on the analysis results. For example, if a user tends to frequently communicate with people who share certain hobbies or preferences, the AI ​​will learn that tendency and recommend people with similar hobbies and preferences as potential matches.

[0040] "Example 3"

[0041] Step 1: AI monitors the matching status of all users.

[0042] Step 2: For users with few matches, the AI ​​will more proactively recommend potential matches. For example, if a user has not had a match for a certain period of time, the AI ​​will proactively recommend users who are likely to match the user's attributes.

[0043] Example 1

[0044] 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."

[0045] In conventional systems, users' profile pictures were based on self-reporting, and often did not fully reflect the user's characteristics and preferences. Furthermore, it was difficult to consider the user's latent attributes and behavioral history when matching, resulting in low matching accuracy. Furthermore, there was also the issue of matching being biased toward certain users, resulting in an imbalance in the overall amount of matches.

[0046] 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.

[0047] In this invention, the server includes means for receiving profile information entered by a user, means for generating a prompt sentence based on the received profile information, means for inputting the generated prompt sentence into a generative AI model and generating a profile picture, means for sending the generated profile picture to the user's device, means for guiding potential matches through advanced attribute judgment by AI, and means for reducing bias in the amount of matches. This makes it possible to generate profile pictures that reflect the user's characteristics and preferences, improving matching accuracy and balancing the overall amount of matches.

[0048] The "means for receiving profile information entered by the user" is a function that allows the server to receive information such as a self-introduction, hobbies, and preferences entered by the user through the terminal.

[0049] "Means for generating a prompt sentence based on received profile information" refers to a function that creates a prompt sentence to be input into the generative AI model based on the user's profile information received by the server.

[0050] The "means for inputting the generated prompt sentence into the generative AI model and generating a profile picture" is a function for inputting the prompt sentence generated by the server into the generative AI model and, as a result, generating a profile picture that matches the user's characteristics and preferences. The server may include a creation means for inputting a specific prompt sentence into the generative AI model, thereby creating a profile picture for the user based on the user's behavior history and profile information.

[0051] "Means for sending the generated profile picture to the user's terminal" is a function that enables the server to send the profile picture obtained from the generative AI model to the user's terminal.

[0052] "Means for guiding potential matches through advanced attribute judgment by AI" refers to a function in which AI analyzes a user's behavioral history and profile information, judges the user's potential attributes, and recommends optimal matches. The server may include means for using a generative AI model to recommend matches for the user based on emotions recognized by the emotion engine.

[0053] "Means for reducing bias in the amount of matching" is a function for balancing the overall amount of matching so that matching is not biased towards specific users.

[0054] MODE FOR CARRYING OUT THE INVENTION

[0055] The present invention provides a system for generating a profile picture based on a user's characteristics and preferences, thereby improving the accuracy of matching. Specific embodiments of this system will be described below.

[0056] Program Generation

[0057] The server generates a program for generating a profile picture for a user. The program uses a generative AI model to create the profile picture based on information provided by the user. Examples of generative AI models that can be used include DALL-E and MidJourney from OpenAI (registered trademark). The server may include a generating unit that generates a profile picture for the user based on the user's behavioral history and profile information by inputting a specific prompt sentence into the generative AI model.

[0058] Program processing

[0059] The server receives profile information entered by the user through the terminal. This profile information includes a self-introduction, hobbies, preferences, etc. For example, if a user enters "I like the outdoors," this information is sent to the server.

[0060] Next, the server generates a prompt based on the received profile information. The prompt is text to be input into the generative AI model, and reflects the user's characteristics and preferences. For example, a prompt might be generated that reads, "Please generate a profile picture for a user who loves the outdoors. Please set the background to an outdoor scene, such as a campsite or mountain."

[0061] The server inputs the generated prompt sentence into the generative AI model, which then generates a profile picture that matches the user's characteristics and preferences based on the prompt sentence. The generated profile picture is then sent to the server.

[0062] The server sends the generated profile picture to the user's device, and the user can check the received profile picture on the device and use it as their profile.

[0063] Specific examples

[0064] For example, if a user enters the information "I like reading," the following prompt will be generated:

[0065] "Generate a profile picture for a user who likes reading. Set the background to a reading scene, such as a library or bookshelf."

[0066] Based on this prompt, the generative AI model generates a profile picture with a library and bookshelves as the background, and the generated picture is sent from the server to the user's device, where the user can use the picture as their profile picture.

[0067] In this way, by generating a profile picture based on the user's characteristics and preferences, the accuracy of matching can be improved. Furthermore, advanced attribute judgment using AI can analyze the user's behavioral history and profile information to recommend the most suitable match. Furthermore, by including a means to reduce bias in the amount of matches, the overall amount of matches can be balanced.

[0068] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0069] Step 1:

[0070] The user enters profile information.

[0071] The user uses the terminal to input profile information such as a self-introduction, hobbies, preferences, etc. For example, the user may input "I like the outdoors." This input information is sent to the server.

[0072] Step 2:

[0073] The server receives the user's input information.

[0074] The server receives the profile information sent by the user from the terminal. This information is text data that reflects the user's hobbies and preferences. The received data is used in the next processing step.

[0075] Step 3:

[0076] The server generates a prompt sentence based on the received profile information.

[0077] The server creates a prompt to input to the generative AI model based on the received user profile information. For example, it generates a prompt such as, "Please generate a profile picture for a user who likes the outdoors. Please set the background to an outdoor scene, such as a campsite or mountain." This prompt is used in the next step.

[0078] Step 4:

[0079] The prompt sentence generated by the server is input into the generative AI model.

[0080] The server inputs the generated prompt into a generative AI model (such as OpenAI's DALL-E or MidJourney), which then generates a profile picture that matches the user's characteristics and preferences based on the prompt. The generative AI model then analyzes the prompt and generates an image with an appropriate background and scene.

[0081] Step 5:

[0082] The server receives the generated profile photo.

[0083] The server receives a profile picture generated from the generative AI model, which includes a background and scene that matches the user's interests and preferences. The received profile picture is used in the next step.

[0084] Step 6:

[0085] The server sends the profile picture to the user's device.

[0086] The server sends the received profile picture to the user's device, where the user can use the picture as their profile. The sent picture is displayed on the user's device.

[0087] Step 7:

[0088] The user checks their profile picture.

[0089] The user can review the received profile photo on their device, check if the photo meets their expectations, and request a regeneration if necessary.

[0090] (Application example 1)

[0091] 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."

[0092] In conventional systems, users' profile pictures are based on self-reporting and do not reflect their hobbies or preferences, which means that they cannot fully express the user's personality and interests. Another issue is that poor quality profile pictures result in a lower matching success rate. Furthermore, because the system does not fully utilize the user's behavioral history and profile information, it is difficult to recommend optimal matches.

[0093] 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.

[0094] In this invention, the server includes a means for creating a profile picture using AI, a means for guiding potential matches through advanced attribute judgment by AI, a means for reducing bias in the amount of matches, a means for generating a prompt sentence based on the user's hobbies and preferences and generating a profile picture using a generative AI model, and a means for reflecting the generated profile picture in the user's profile. This makes it possible to generate a high-quality profile picture that reflects the user's personality and interests and recommend optimal matches. The server may include a creation means for creating a profile picture for the user based on the user's behavioral history and profile information by inputting a specific prompt sentence into the generative AI model. The server may also include a means for recommending matches for the user based on emotions recognized by an emotion engine using the generative AI model.

[0095] An "AI profile picture generation method" is a method that uses artificial intelligence technology to generate a profile picture based on user-provided characteristics and preferences.

[0096] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match.

[0097] "Means for inducing potential matches" are means for finding potentially compatible partners based on the user's attributes and behavioral history, and promoting matching.

[0098] "Means for reducing bias in the amount of matching" refers to means for evenly distributing the overall amount of matching so that matching is not concentrated on specific users.

[0099] "Means for generating prompt sentences based on the user's hobbies and preferences" refers to means for creating prompt sentences to be input into a generative AI model based on information about hobbies and preferences provided by the user.

[0100] "Means for generating a profile picture using a generative AI model" refers to means for generating a profile picture for a user using a generative AI model based on a generated prompt sentence.

[0101] The "means for reflecting the generated profile picture in the user's profile" refers to a means for adding the generated profile picture to the user's profile information and displaying it.

[0102] The system for implementing this invention is composed of three main elements: a server, a terminal, and a user. The specific operations of each element and their cooperation will be explained below.

[0103] server

[0104] The server includes a means for creating a profile picture using AI, a means for determining advanced attributes using AI, a means for guiding potential matches, a means for reducing bias in the amount of matches, a means for generating prompt sentences based on the user's hobbies and preferences, a means for generating a profile picture using a generative AI model, and a means for reflecting the generated profile picture in the user's profile.

[0105] The server receives the user's hobbies and preferences and generates a prompt based on them. Using the generated prompt, it uses a generative AI model (e.g., OpenAI's API) to generate a profile picture for the user. The generated profile picture is added to the user's profile information and displayed.

[0106] Terminal

[0107] The terminal is a device where the user enters information and checks the generated profile picture. It can be a smartphone or a PC. The user enters information about their hobbies and preferences through the terminal and sends it to the server. The generated profile picture can be checked on the terminal.

[0108] User

[0109] A user is an individual who uses the system. The user inputs information about their hobbies and preferences through their device and sends it to the server. The server generates a profile picture, which the user can then view on their device.

[0110] Specific examples

[0111] For example, if the user provides information such as "I like fashion," the server generates the following prompt sentence:

[0112] Prompt Sentence Examples

[0113] The user's hobby is fashion, and their preference is stylish cityscapes.

[0114] Based on this prompt, a generative AI model is used to generate a profile picture featuring a fashion show or stylish cityscape as a background, which is then added to the user's profile information and can be viewed on the device.

[0115] In this way, it is possible to generate high-quality profile pictures that reflect the user's personality and interests, and to recommend optimal matches.

[0116] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0117] Step 1:

[0118] The user uses the terminal to input information about his or her hobbies and preferences.

[0119] Input: User's hobbies and preferences (e.g., "I like fashion")

[0120] Output: The entered information is sent to the server.

[0121] Specific operation: The user enters information about their hobbies and preferences into the input form on the device and presses the "Send" button. The device then sends this information to the server.

[0122] Step 2:

[0123] The server generates a prompt based on the received user information.

[0124] Input: User's hobbies and preferences

[0125] Output: Generated prompt sentence (e.g., "The user's hobby is fashion, and their preference is stylish cityscapes.")

[0126] Specific operation: The server analyzes the received information and generates a prompt sentence based on the template.

[0127] Step 3:

[0128] The server sends a request to the generative AI model using the generated prompt sentence.

[0129] Input: Generated prompt text

[0130] Output: Profile photo URL returned by the generative AI model

[0131] Specific operation: The server sends the generated prompt to a generative AI model (e.g., OpenAI's API) and requests the generation of a profile photo. The generative AI model generates an image based on the prompt and returns its URL to the server.

[0132] Step 4:

[0133] The server receives the profile picture URL returned by the generative AI model and downloads the image.

[0134] Input: Profile photo URL

[0135] Output: Downloaded profile photo

[0136] What happens: The server uses the returned URL to download the profile picture.

[0137] Step 5:

[0138] The server adds the downloaded profile picture to the user's profile information and sends it to the terminal.

[0139] Input: Downloaded profile photo

[0140] Output: Updated user profile information

[0141] Specific operation: The server adds the downloaded profile picture to the user's profile information and sends the updated profile information to the terminal.

[0142] Step 6:

[0143] The user uses the device to view the generated profile picture.

[0144] Input: Updated user profile information

[0145] Output: The generated profile picture displayed on the device.

[0146] Specific operation: The user checks the generated profile picture through the device, and the device displays the updated profile information received from the server.

[0147] Example 2

[0148] 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."

[0149] Conventional systems have the problem of being unable to fully utilize users' behavioral history and profile information, resulting in low accuracy in recommending optimal matches. Furthermore, insufficient data preprocessing reduces the learning efficiency of the AI ​​model, resulting in poor quality recommendation results.

[0150] 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.

[0151] In this invention, the server includes means for collecting user behavior history and profile information, means for preprocessing the collected data, means for training an AI model using the preprocessed data, and means for recommending optimal matching partners using the trained AI model. This makes it possible to effectively utilize the user's behavior history and profile information and recommend optimal matching partners with high accuracy. The server may also include means for using the generative AI model to recommend matching partners for the user based on emotions recognized by the emotion engine.

[0152] "User behavior history" is a record of the operations and activities that a user performs on the system.

[0153] "Profile information" refers to personal information and attribute information provided by a user at the time of registration.

[0154] "Means of collection" refers to the methods and technologies used to store user behavior history and profile information in a database.

[0155] "Preprocessing means" refers to methods and techniques that process collected data through cleaning, normalization, feature engineering, etc., to convert it into a format that is easy for an AI model to learn.

[0156] "Means for training an AI model" refers to methods or techniques that use preprocessed data to apply machine learning algorithms to learn user behavior patterns and attributes.

[0157] "Means for recommending optimal matching partners" refers to methods and technologies that use trained AI models to predict and present optimal matching partners to users.

[0158] This invention is a system that recommends optimal matches by collecting user behavior history and profile information and generating an AI model based on this information. Specific embodiments of this system are described below.

[0159] First, the server collects the user's behavioral history and profile information. The behavioral history includes the actions the user has taken on the website or application (e.g., page viewing, message sending, likes). The profile information includes the personal information and attribute information (e.g., age, gender, hobbies, occupation) entered by the user when registering. This data is stored on a database server (e.g., MySQL (registered trademark), PostgreSQL). A data collection tool (e.g., Apache (registered trademark) Kafka) is used to collect the data.

[0160] Next, the server preprocesses the collected data. Preprocessing includes data cleaning (e.g., imputing missing values ​​and removing outliers), normalization (e.g., scaling numerical data), and feature engineering (e.g., generating new features). This converts the data into a format suitable for training AI models. Data preprocessing tools (e.g., Pandas and NumPy) are used for data preprocessing.

[0161] The server then uses the preprocessed data to train an AI model using a machine learning framework (e.g., TENSORFLOW (registered trademark), PyTorch). The AI ​​model learns the user's behavioral patterns and attributes to predict the best match.

[0162] Finally, the server uses the trained AI model to recommend the best possible match for the user. Specifically, the user's behavioral history and profile information are input into the AI ​​model, and the model presents the best possible match to the user. The recommendation system uses recommendation algorithms (e.g., Scikit-learn, LightGBM).

[0163] Examples:

[0164] If user A has hobbies of "hiking" and "cooking" and frequently communicates with other users who have these hobbies, the AI ​​will learn this tendency. Next, if user B also has hobbies of "hiking" and "cooking," the AI ​​will recommend user A and user B as potential matches.

[0165] Example prompt sentence:

[0166] "Generate an AI model that recommends optimal matches based on a user's behavioral history and profile information. User A's hobbies are hiking and cooking, and he frequently communicates with other users who share these hobbies. User B's hobbies are also hiking and cooking. Based on this information, create an AI model that recommends User A and User B as potential matches."

[0167] The above is a specific embodiment for carrying out the present invention.

[0168] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0169] Step 1:

[0170] Data collection

[0171] The server collects the user's behavioral history and profile information.

[0172] Input: User operation history (e.g., page viewing, message sending, likes) and registration information (e.g., age, gender, hobbies, occupation)

[0173] Specific operations: The server logs the actions users take on websites and applications and stores them in a database. Data collection tools (e.g., Apache Kafka) are used to collect data in real time.

[0174] Output: User behavior history and profile information stored in a database

[0175] Step 2:

[0176] Data Preprocessing

[0177] The server pre-processes the collected data.

[0178] Input: User behavior and profile information stored in a database

[0179] Specific operations: The server performs data cleaning (e.g., missing value imputation, outlier removal), normalization (e.g., scaling numerical data), and feature engineering (e.g., generating new features) using data preprocessing tools (e.g., Pandas, NumPy).

[0180] Output: Preprocessed dataset

[0181] Step 3:

[0182] Training an AI model

[0183] The server uses the preprocessed data to train the AI ​​model.

[0184] Input: Preprocessed dataset

[0185] Specific operation: The server uses machine learning algorithms (e.g., deep learning, gradient boosting) to learn user behavior patterns and attributes. It uses machine learning frameworks (e.g., TensorFlow, PyTorch).

[0186] Output: A trained AI model

[0187] Step 4:

[0188] Matching partner recommendations

[0189] The server uses a trained AI model to recommend the best possible match for the user.

[0190] Input: Trained AI model, user behavior history and profile information

[0191] Specific operation: The server inputs the user's behavioral history and profile information into the AI ​​model, and presents the user with the best match predicted by the model. It uses a recommendation system (e.g., Scikit-learn, LightGBM).

[0192] Output: A list of best matches presented to the user

[0193] The above is the specific processing flow of the program of this system.

[0194] (Application example 2)

[0195] 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."

[0196] Conventional systems were unable to fully utilize a user's behavioral history and profile information, making it difficult to recommend optimal matches and products. Furthermore, there was a tendency for matching volume to be biased, reducing user satisfaction. Furthermore, it was difficult to make highly accurate product recommendations based on a user's interests and preferences.

[0197] 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.

[0198] In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches using advanced attribute judgments using AI, means for reducing bias in the amount of matches, means for analyzing the user's behavioral history and profile information to recommend optimal products, and means for vectorizing product descriptions and calculating cosine similarity to recommend similar products. This makes it possible to recommend optimal matches and products with high accuracy based on the user's behavioral history and profile information.

[0199] "AI-based profile picture creation method" refers to a method that uses artificial intelligence technology to generate a profile picture based on a user's characteristics and preferences.

[0200] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match and products.

[0201] "Means for inducing potential matches" refers to means for finding and recommending potential matches based on the user's attributes and behavioral history.

[0202] "Means for reducing bias in the amount of matching" refers to means for reducing bias in matching for particular users and providing equal opportunities for matching.

[0203] "Means for analyzing a user's behavioral history and profile information and recommending optimal products" refers to means for recommending optimal products to a user based on the user's past behavior and profile information.

[0204] "Means for vectorizing product descriptions, calculating cosine similarity, and recommending similar products" refers to a means for finding and recommending similar products by converting product descriptions into numerical vectors and calculating cosine similarity.

[0205] A system for implementing this invention includes a server, a user terminal, and an AI model. The server collects user behavioral history and profile information and runs software to analyze this data. Specifically, the server processes the data using libraries such as Python, Pandas, and Scikit-learn.

[0206] The server first loads the user's behavioral history data and product data. Next, it uses TfidfVectorizer to vectorize the product descriptions and extract the user's interests. This allows it to obtain vectors of products that the user has shown interest in in the past. It then calculates the cosine similarity and recommends the most similar products.

[0207] The user terminal is a device such as a smartphone or PC that displays the recommendation results sent from the server. The user can then check and purchase the recommended products.

[0208] For example, if a user frequently purchases "sports goods," the server will recommend the latest sports goods and related products based on that user's behavioral history, allowing the user to easily find products that match their interests.

[0209] An example of a prompt is as follows:

[0210] "Create a Python program that recommends optimal products based on user behavior history data. Include a function to extract user interests and calculate cosine similarity to recommend the most similar products."

[0211] In this way, the system for implementing the invention is able to recommend optimal matches and products with a high degree of accuracy based on the user's behavioral history and profile information.

[0212] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0213] Step 1:

[0214] The server reads user behavior history data and product data. Specifically, it obtains data from files named user_behavior.csv and product_catalog.csv. The input is user behavior history data and product data, and the output is a data frame containing these data. The server reads these data using the Pandas library.

[0215] Step 2:

[0216] The server vectorizes the product description. Specifically, it converts the product description into a numeric vector using TfidfVectorizer. The input is the product description text, and the output is the vectorized product description. The server performs this process using the Scikit-learn library.

[0217] Step 3:

[0218] The server obtains vectors of products of interest from the user's behavioral history. Specifically, it extracts corresponding vectors based on the IDs of products that the user has previously shown interest in. The input is the user's behavioral history data and vectorized product descriptions, and the output is a vector that reflects the user's interests.

[0219] Step 4:

[0220] The server calculates the cosine similarity between the vector reflecting the user's interests and the vector of all products. The input is the vector reflecting the user's interests and the vector of all products, and the output is the similarity score. The server performs this calculation using the Scikit-learn library.

[0221] Step 5:

[0222] The server recommends the most similar products. Specifically, it selects the product with the highest score based on the cosine similarity score and recommends it to the user. The input is the similarity score, and the output is a list of recommended products.

[0223] Step 6:

[0224] The terminal displays the recommendation results sent from the server. Specifically, it displays a list of recommended products on the user's smartphone or PC. The input is the recommendation results sent from the server, and the output is the product list displayed to the user.

[0225] Step 7:

[0226] The user checks the recommended products and purchases them. Specifically, they select products of interest from the product list displayed on the terminal and then carry out the purchase procedure. The input is the list of recommended products, and the output is information about the purchased product.

[0227] Example 3

[0228] 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."

[0229] In conventional systems, there was a problem that the amount of matching between users was easily biased, and some users did not get matching opportunities. In addition, there was an issue that appropriate recommendations were not made to users with few matches, resulting in a poor user experience.

[0230] 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.

[0231] In this invention, the server includes means for storing the matching status of users in a database, means for the server to periodically acquire the matching status of all users, means for the server to input the matching status to the AI ​​model, means for the AI ​​model to identify users with few matches, means for the AI ​​model to select recommended users for the identified users, and means for the server to notify target users of information about the recommended users. This makes it possible to reduce bias in the amount of matches and make appropriate recommendations to users with few matches.

[0232] "Means for saving a user's matching status in a database" is a function that records information about a new match made by a user in a database.

[0233] "Means for the server to periodically obtain the matching status of all users" refers to a function by which the server obtains the matching status of all users from the database at regular time intervals.

[0234] "Means for the server to input the matching status into the AI ​​model" refers to the function by which the server inputs the matching status data it has acquired into the AI ​​model.

[0235] The "means for the AI ​​model to identify users with few matches" is a function in which the AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time. The server may include means for monitoring the matching status of all users and recommending matching partners to users whose number of matches is fewer than the number of matches of a specific user.

[0236] "Means for selecting recommended users for a user identified by an AI model" is a function that selects recommended users for a user identified by an AI model based on attributes and past matching history.

[0237] "Means for the server to notify the target user of information about recommended users" is a function by which the server notifies the target user of information about recommended users selected by the AI ​​model.

[0238] The "scheduler" is a function that allows the server to automate the process of periodically obtaining the matching status of all users.

[0239] "User attributes" refers to personal information such as the user's age, hobbies, and region.

[0240] MODE FOR CARRYING OUT THE INVENTION

[0241] The present invention relates to a system that reduces bias in the amount of matching for users and makes appropriate recommendations to users with few matches. Specific embodiments of this system will be described below.

[0242] System configuration

[0243] This system consists of three main elements: a server, a terminal, and a user. The server contains software components such as a database, an AI model, and a scheduler. A terminal is a device used by a user (such as a smartphone, tablet, or PC). A user is an individual who uses the system.

[0244] Hardware and software used

[0245] Server: Use a high-performance server machine (e.g., AWS (registered trademark) EC2 instance). Use MySQL as the database and TensorFlow or PyTorch to build the AI ​​model.

[0246] Device: A user device such as a smartphone, tablet, or PC.

[0247] Scheduler: Use scheduling tools such as cron jobs to automate periodic data retrieval.

[0248] Program processing

[0249] The server stores the matching status of users in a database. When a user makes a new match, the device records that information in the database. The server retrieves the matching status of all users from the database at regular intervals (e.g., every hour). This process is automated using a scheduler.

[0250] The server inputs the acquired matching status data into the AI ​​model. The AI ​​model performs analysis based on the user's attributes (age, hobbies, region, etc.) and past matching history. The AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time (e.g., one week). These identified users are listed as users with few matches.

[0251] The AI ​​model recommends other users who are likely to have matching attributes and hobbies to the identified user. The recommendation algorithm is based on the user's attributes, such as age, hobbies, and region. The server notifies the target user of information about the recommended users selected by the AI ​​model. Notifications are sent to the device via push notification or email. By receiving the notification, users gain new matching opportunities.

[0252] Specific examples

[0253] For example, if User A (age 30, hobby: hiking, location: Tokyo) has not had a match for more than a week, the server retrieves User A's matching status from the database. The server inputs this data into the AI ​​model, which then recommends User B (age 28, hobby: hiking, location: Tokyo), who is likely to be a good match for User A. The server then sends User A a push notification with User B's profile. User A receives the notification and has the opportunity to newly match with User B.

[0254] Prompt Sentence Examples

[0255] "If User A, age 30, has not had a match for more than a week, the AI ​​should recommend User B, age 28, who enjoys hiking."

[0256] In this way, the system processing can be performed while clarifying the roles of the server, terminal, and user. The flow of the specific processing in the third embodiment will be described with reference to FIG.

[0257] Step 1:

[0258] When a user makes a new match, the device stores that information in a database.

[0259] Input: User's matching information (user ID, matched person's ID, matching date and time)

[0260] Output: Matching information stored in a database

[0261] Specific operation: When a user matches, the device records the user ID, the matched person's ID, and the date and time of the match in a database.

[0262] Step 2:

[0263] The server periodically obtains the matching status of all users.

[0264] Input: All user matching information stored in the database

[0265] Output: Matching status data for all users obtained by the server

[0266] Specific operation: The server uses a scheduler to retrieve the matching status of all users from the database at regular intervals (e.g., every hour).

[0267] Step 3:

[0268] The matching status data acquired by the server is input into the AI ​​model.

[0269] Input: Matching status data of all users acquired by the server

[0270] Output: Matching situation data input into the AI ​​model

[0271] Specific operation: The server converts the acquired matching status data into an appropriate format and passes it to the AI ​​model in order to input it into the AI ​​model.

[0272] Step 4:

[0273] The AI ​​model identifies users with low matches.

[0274] Input: Matching situation data input into the AI ​​model

[0275] Output: List of users with few matches

[0276] Specific operation: The AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time (e.g., one week).

[0277] Step 5:

[0278] The AI ​​model selects recommended users for the identified user.

[0279] Input: List of users with few matches, user attribute data

[0280] Output: A list of recommended users

[0281] Specific operation: The AI ​​model selects recommended users for the identified user based on their attributes and past matching history.

[0282] Step 6:

[0283] The server notifies the target user of the information about the recommended user.

[0284] Input: List of recommended users

[0285] Output: Information about the recommended user notified to the target user

[0286] Specific operation: The server notifies the target user of the information about the recommended users selected by the AI ​​model. Notifications are sent to the device via push notification or email.

[0287] (Application example 3)

[0288] 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."

[0289] Conventional systems tend to be prone to bias in the amount of matching a user receives, and there was a problem in that appropriate responses were not made to users who had not been matched for a certain period of time. Furthermore, even on online shopping sites, it was difficult to effectively recommend specific products because users' purchase and browsing histories could not be fully utilized. This led to issues such as lower user satisfaction and a decrease in the frequency of site use.

[0290] 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.

[0291] In this invention, the server includes a means for creating a profile picture using AI, a means for inducing potential matches using advanced attribute judgment by AI, a means for reducing bias in the amount of matches, a means for analyzing a user's purchase history and browsing history and recommending specific products, and a means for actively recommending products to users who have not made a purchase for a certain period of time. This makes it possible to reduce bias in the amount of matches for users and effectively recommend specific products.

[0292] A "means for creating a profile picture" is a means for using AI to generate a profile picture based on a user's characteristics and preferences.

[0293] "Advanced attribute judgment" means that AI analyzes a user's behavioral history and profile information to make judgments to recommend the most suitable match.

[0294] The "means for reducing the bias in the number of matches" refers to a means for an AI to monitor the matching status of all users and actively recommend matching partners to users with few matches. The server may include a means for monitoring the matching status of all users and recommending matching partners to users with fewer matches than the number of matches of a specific user.

[0295] "Means for analyzing purchase history and browsing history" refers to a means by which AI analyzes a user's past purchase history and browsing history to understand their interests and concerns.

[0296] "Means of recommending specific products" refers to a means of recommending products that match the user based on the user's purchasing history and browsing history analyzed by AI.

[0297] The "means for actively recommending products to users who have not made a purchase for a certain period of time" refers to a means for identifying users who have not made a purchase for a certain period of time and actively recommending products to those users.

[0298] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The server includes a means for creating profile pictures using AI, a means for determining advanced attributes, a means for reducing bias in the amount of matching, a means for analyzing purchase history and browsing history, a means for recommending specific products, and a means for proactively recommending products to users who have not made a purchase for a certain period of time.

[0299] Server Processing

[0300] The server first collects user profile information and behavioral history, then analyzes this data using AI. Specifically, it uses the pandas library to load user and product data, and uses scikit-learn's cosine_similarity to recommend products based on the user's attributes. This allows it to generate a profile picture based on the user's characteristics and preferences, and recommend optimal matches and products.

[0301] Terminal handling

[0302] The device provides an interface for users to access the service. Through a device such as a smartphone or PC, users can create a profile picture and check the results of product recommendations. In addition, if a user has not made a purchase for a certain period of time, the system will actively recommend products to them via push notifications and emails.

[0303] User operations

[0304] Users enter their profile information through their devices and provide their purchase and browsing history. This allows the server to accurately understand the user's attributes and recommend the most suitable match and products. The user can then check the recommended products and purchase them.

[0305] Specific examples

[0306] For example, if a female user in her 30s is interested in fashion and has previously purchased casual wear and sneakers, the server will analyze the user's purchasing history and recommend the following products:

[0307] Example prompt sentence:

[0308] User ID: 12345

[0309] Attributes: Women in their 30s, fashion enthusiasts, mid-range price range

[0310] Purchase history: casual wear, sneakers, accessories

[0311] Recommended products:

[0312] casual jacket

[0313] Sports shoes

[0314] Simple necklace

[0315] By inputting this prompt sentence into a generative AI model, the most suitable product recommendations are made to the user.

[0316] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0317] Step 1:

[0318] The server reads user data and product data. Specifically, it reads data from a CSV file using the pandas library. The input is the CSV file of user data and product data, and the output is the user data and product data in data frame format.

[0319] Step 2:

[0320] The server analyzes the user's purchase history and browsing history. Specifically, it extracts the purchase history of a specific user and obtains the user's attribute information. The input is the user ID, and the output is the user's purchase history data.

[0321] Step 3:

[0322] The server uses an AI model to recommend products based on user attributes. Specifically, it uses scikit-learn's cosine_similarity to calculate the similarity between the user's attribute vector and the product vector. The input is the user's attribute information and product data, and the output is a list of products with high similarity.

[0323] Step 4:

[0324] The server identifies users who have not made a purchase for a certain period of time. Specifically, it compares the current date with the user's last purchase date and extracts users who have not made a purchase for a certain period of time (e.g., 30 days). The input is user data, and the output is a list of users who have not made a purchase for a certain period of time.

[0325] Step 5:

[0326] The server proactively recommends products to users who have not made a purchase for a certain period of time. Specifically, it uses the product list obtained in step 3 to send push notifications and emails to users. The input is a list of users who have not made a purchase for a certain period of time and a product list, and the output is a notification of recommended products sent to the user.

[0327] Step 6:

[0328] The terminal provides an interface that allows users to check recommended products. Specifically, it displays recommended products on the screen of a smartphone or PC. The input is recommended product data sent from the server, and the output is recommended product information that is displayed to the user.

[0329] Step 7:

[0330] The user checks the recommended products through the terminal and purchases them. Specifically, they click on the recommended products to check detailed information and then complete the purchase procedure. The input is the recommended product information displayed on the terminal, and the output is the completion of the purchase procedure.

[0331] 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.

[0332] "Example 1"

[0333] In one embodiment of the present invention, in addition to a means for creating a profile picture using AI, a means for determining advanced attributes using AI, and a means for reducing bias in the amount of matches, an emotion engine that recognizes the user's emotions is incorporated. This emotion engine recognizes emotions from the user's text input or voice input and recommends matching partners based on those emotions. For example, if a user inputs the text "I'm having fun," the emotion engine recognizes that the user is currently in a happy mood and recommends matching partners that match that mood.

[0334] "Example 2"

[0335] In another embodiment of the present invention, the emotion engine tracks changes in the user's emotions and re-recommends a match based on those changes. For example, if a user first inputs the text "fun" and then later inputs the text "lonely," the emotion engine recognizes that the user's emotion has changed from fun to loneliness and re-recommends a match based on that new emotion.

[0336] Re-recommend matching partners that match your emotions.

[0337] "Example 3"

[0338] Furthermore, in another embodiment of the present invention, the emotion engine learns changes in the user's emotions and recommends matching partners based on the learning results. For example, if a user tends to input the text "fun" every weekend, the emotion engine learns this tendency and recommends matching partners that match the user's happy mood on weekends.

[0339] The processing flow of each embodiment will be described below.

[0340] "Example 1"

[0341] Step 1: The user enters text or voice input into the application.

[0342] Step 2: The emotion engine recognizes emotions from the user's input. For example, if the text "fun" is entered, the emotion engine recognizes that the user is in a happy mood.

[0343] Step 3: Based on the emotions recognized by the emotion engine, the AI ​​recommends potential matches, taking into account the user's profile information and past behavioral history.

[0344] "Example 2"

[0345] Step 1: The user enters text or voice input into the application.

[0346] Step 2: The emotion engine recognizes emotions from the user's input and tracks changes in the emotions. For example, if the user first inputs "happy" and then later inputs "lonely," the emotion engine recognizes the change in emotion. The server may include a means for tracking changes in the user's emotions.

[0347] Step 3: Based on the changes in emotions recognized by the emotion engine, the AI ​​re-recommends potential matches, taking into account the user's profile information and past behavioral history.

[0348] "Example 3"

[0349] Step 1: The user enters text or voice input into the application.

[0350] Step 2: The emotion engine recognizes emotions from the user's input and learns changes in those emotions. For example, if a user tends to input "fun" every weekend, the emotion engine will learn that tendency.

[0351] Step 3: Based on the emotional trends learned by the emotion engine, the AI ​​recommends potential matches, taking into account the user's profile information and past behavioral history.

[0352] Example 1

[0353] 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."

[0354] In conventional systems, it was difficult to fully reflect the user's characteristics and emotions when generating a user's profile picture or recommending potential matches. Furthermore, since matchmaking recommendations were not based on the user's emotions, there was a risk of user satisfaction decreasing. This led to a need for improved matching accuracy and user experience.

[0355] 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.

[0356] In this invention, the server includes means for inputting user profile information, means for receiving the input information, means for inputting a prompt sentence to the generative AI model, means for the generative AI model to generate a profile picture, means for analyzing the user's emotions using an emotion engine, means for recommending matching partners based on the emotions, and means for the user to confirm the generated profile picture and the recommended matching partners. This makes it possible to generate a profile picture that reflects the user's characteristics and emotions, and to recommend optimal matching partners based on the emotions.

[0357] The "means for inputting user profile information" refers to a device or software that provides an interface for users to input information such as self-introduction, hobbies, and preferences.

[0358] "Means for receiving input information" refers to a device or software that has the functionality to allow the server to receive, store, or process the profile information input by the user.

[0359] "Means for inputting prompt sentences into a generative AI model" refers to a device or software for inputting prompt sentences generated based on a user's profile information into an AI model.

[0360] A "means for a generative AI model to generate a profile picture" is a device or software that uses AI technology to generate a profile picture based on a user's characteristics and preferences.

[0361] The "means for analyzing user emotions using an emotion engine" is a device or software for analyzing emotions from a user's text input or voice input.

[0362] The "means for recommending a match based on emotion" is a device or software for recommending the most suitable match based on the analyzed emotion information of the user.

[0363] The "means for the user to review the generated profile picture and recommended matches" refers to a device or software that provides an interface that allows the user to review the generated profile picture and recommended matches.

[0364] MODE FOR CARRYING OUT THE INVENTION

[0365] This invention is a system that generates a profile picture using a generative AI model based on a user's profile information, analyzes the user's emotions using an emotion engine, and recommends optimal matches. Specific embodiments of this system are described below.

[0366] Program Generation

[0367] The server creates a program to generate a user's profile picture. This program uses a generative AI model to generate a profile picture based on information such as a self-introduction, hobbies, and preferences provided by the user. It also includes a function to analyze the user's emotions using an emotion engine and recommend potential matches based on those emotions.

[0368] Program processing

[0369] The server receives profile information entered by the user through the device. For example, the user might enter information such as "I like the outdoors and often go camping." Based on this information, the server generates a prompt sentence and inputs it into the generative AI model. Image generation models such as "Stable Diffusion" and "DALL-E" can be used as the generative AI model.

[0370] The generative AI model generates a profile picture based on a prompt received from the server. For example, the following prompt is input to the generative AI model:

[0371] The user's hobby is the outdoors, and they like camping. Based on this information, generate a profile picture with an outdoor scene in the background, such as a campsite or a mountain.

[0372] Based on this prompt, the generative AI model generates a profile photo with a campsite as the background.

[0373] Furthermore, the server uses an emotion engine to analyze the user's emotions. If the user inputs "I'm in a happy mood today," the emotion engine analyzes the information and recognizes that the user is in a happy mood. The emotion engine can use text analysis or voice analysis technology.

[0374] The server then recommends a match based on the user's emotional information. For example, if a user is in a happy mood, the server recommends users who are in a similar happy mood or who share fun activities.

[0375] Specific examples

[0376] If the user inputs "I like the outdoors and often go camping," the server will input the following prompt sentence into the generative AI model:

[0377] The user's hobby is the outdoors, and they like camping. Based on this information, generate a profile picture with an outdoor scene in the background, such as a campsite or a mountain.

[0378] Based on this prompt, the generative AI model generates a profile photo with a campsite as the background.

[0379] For example, if a user types "I'm in a happy mood today," the server will use its emotion engine to analyze the user's emotions and recommend a match that matches their happy mood. For example, it will recommend users who also like the outdoors and are in a happy mood.

[0380] In this way, the server can use AI technology based on the information provided by the user to generate profile pictures and recommend potential matches.

[0381] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0382] Program processing flow

[0383] Step 1: User enters profile information

[0384] The user inputs profile information such as a self-introduction, hobbies, and preferences through the terminal. For example, the user may input information such as "I like the outdoors and often go camping."

[0385] Input: User profile information

[0386] Output: Sending profile information from device to server

[0387] Step 2: The server receives the entered information

[0388] The server receives the profile information sent by the user from the terminal, which is used in subsequent processing.

[0389] Input: Profile information sent from the device

[0390] Output: Profile information stored on the server

[0391] Step 3: The server inputs a prompt to the generative AI model.

[0392] The server generates a prompt based on the received profile information and inputs it into the generative AI model. For example, it generates a prompt such as, "The user's hobby is the outdoors, and they like camping. Based on this information, please generate a profile picture with an outdoor scene, such as a campsite or mountain, as the background."

[0393] Input: Profile information stored on the server

[0394] Output: The prompt sentence that is input to the generative AI model

[0395] Step 4: Generative AI model generates profile photo

[0396] Generative AI models (such as "Stable Diffusion" and "DALL-E") generate profile pictures based on prompts received from the server, with backgrounds that match the user's interests and preferences.

[0397] Input: A prompt sentence entered into the generative AI model

[0398] Output: Generated profile picture

[0399] Step 5: The server uses the emotion engine to analyze the user's emotions.

[0400] The server uses an emotion engine to analyze emotions from text or voice input by the user. For example, if a user inputs "I'm in a happy mood today," the emotion engine analyzes the information and recognizes that the user is in a happy mood.

[0401] Input: User-entered text or speech

[0402] Output: Analyzed user emotion information

[0403] Step 6: The server recommends matches based on emotions

[0404] The server recommends matching partners based on the emotional information analyzed by the emotion engine. For example, if a user is in a happy mood, the server will recommend users who are in a similar happy mood or who can share fun activities.

[0405] Input: Analyzed user emotion information

[0406] Output: A list of suggested matches

[0407] Step 7: User reviews generated profile photo and suggested matches

[0408] Users can view the profile picture generated and the recommended matches through their device, and decide their next action based on this information.

[0409] Input: A generated profile picture and a list of suggested matches

[0410] Output: User confirmation and next action

[0411] (Application example 1)

[0412] 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."

[0413] In conventional systems, the generation of user profile pictures and avatars does not adequately reflect the user's characteristics and preferences, resulting in low user satisfaction. In addition, since matching does not take the user's emotions into consideration, matching accuracy is low, and the user experience is not improved.

[0414] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means. In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches through advanced attribute judgment using AI, means for reducing bias in the amount of matches, means for recognizing a user's emotions and adjusting the facial expression and pose of an avatar based on the emotions, and means for generating an avatar based on the user's characteristics and preferences. This makes it possible to generate a profile picture and avatar based on the user's characteristics and preferences, and further, since matching is performed taking the user's emotions into consideration, it is possible to improve user satisfaction and matching accuracy. The server may also include means for adjusting the facial expression and pose of the user's avatar based on the user's characteristics, preferences, and emotions.

[0415] "Means for creating profile pictures using AI" refers to a means for generating profile pictures using artificial intelligence technology based on information such as a self-introduction, hobbies, and preferences provided by the user.

[0416] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match.

[0417] "Means for inducing potential matches" refers to means for finding and recommending potential matches based on the user's characteristics and preferences.

[0418] "Means for reducing bias in the amount of matching" refers to means for avoiding concentration of matching on specific users and for achieving a balance in overall matching.

[0419] The "means for recognizing a user's emotion and adjusting the avatar's facial expression and pose based on the emotion" refers to a means for recognizing an emotion from a user's text input or voice input and automatically changing the avatar's facial expression and pose based on the emotion. The server may include a means for adjusting the user's avatar's facial expression and pose based on the user's characteristics, the user's preferences, and the emotion.

[0420] "Means for generating an avatar based on a user's characteristics and preferences" refers to a means for using artificial intelligence technology to generate an avatar that is optimal for a user based on information such as a self-introduction, hobbies, and preferences provided by the user.

[0421] The following describes an embodiment of the present invention: First, the overall system configuration will be described, and then the specific operation of each means will be described.

[0422] System configuration

[0423] The system includes a user terminal, a server, and an AI model. The user terminal provides an interface for the user to input information such as a self-introduction, hobbies, and preferences. The server receives this information and generates a profile picture and an avatar using the AI ​​model. The server also recognizes the user's emotions using an emotion engine and adjusts the avatar's facial expression and pose based on the emotions. The server may include means for adjusting the facial expression and pose of the user's avatar based on the user's characteristics, the user's preferences, and the emotions.

[0424] Program processing

[0425] The server receives the information provided by the user and performs the following processes:

[0426] 1. Generate profile photo:

[0427] The server uses a generative AI model to generate a profile picture based on information entered by the user, such as a self-introduction, hobbies, preferences, etc. For example, if a user provides information that they "like the outdoors," the AI ​​model will generate a profile picture with an outdoor scene, such as a campsite or mountain, as the background.

[0428] 2. Avatar generation:

[0429] The server generates an avatar based on the user's characteristics and preferences, and the generated avatar is optimized for the information provided by the user.

[0430] 3. Emotion Recognition and Facial Expression Modulation:

[0431] The server uses an emotion engine that recognizes emotions from the user's text and voice input. For example, if the user inputs the text "I'm having fun," the emotion engine recognizes that the user is currently in a happy mood and adjusts the avatar's facial expression and pose based on that emotion.

[0432] Hardware and software used

[0433] Hardware:

[0434] User devices: smartphones, tablets, computers, etc.

[0435] Server: A server with high-performance computing power

[0436] software:

[0437] Generative AI models: OpenAI APIs, etc.

[0438] Emotion Engine: Emotion recognition engine using natural language processing technology

[0439] Image processing library: PIL (Python Imaging Library) etc.

[0440] Specific examples

[0441] If the user provides information that "I like the outdoors" and also inputs the emotion "fun," the following prompt sentence is generated:

[0442] Prompt Sentence Examples

[0443] Generate an avatar with an outdoor scene background based on the user's bio: I love the outdoors, and my hobbies are camping and hiking.

[0444] By inputting this prompt into a generative AI model, an avatar based on the user's characteristics and preferences is generated. In addition, if the emotion engine recognizes the user's emotion as "happy," the avatar's facial expression is adjusted to a smile.

[0445] In this way, it becomes possible to generate profile pictures and avatars based on the user's characteristics and preferences, and matching is performed taking the user's emotions into consideration, thereby improving user satisfaction and matching accuracy.

[0446] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0447] Step 1:

[0448] The user uses the device to input information such as a self-introduction, hobbies, and preferences. The input information is then sent from the device to the server. Examples of input data include "I like the outdoors," "camping," and "hiking."

[0449] Step 2:

[0450] The server generates a prompt based on the received user information. The generated prompt is then input into the generative AI model. Specifically, the following prompt is generated:

[0451] Generate an avatar with an outdoor scene background based on the user's bio: I love the outdoors, and my hobbies are camping and hiking.

[0452] By inputting this prompt into a generative AI model, an avatar based on the user's characteristics and preferences is generated.

[0453] Step 3:

[0454] The generative AI model generates an avatar based on the prompt text. The image data of the generated avatar is returned to the server. The server receives this image data and proceeds to the next processing step.

[0455] Step 4:

[0456] The server uses an emotion engine to recognize emotions from the user's text input or voice input. For example, if the user inputs the text "fun," the emotion engine recognizes that the user is currently in a happy mood. The input data is the text "fun," and the output data is the emotion information "fun."

[0457] Step 5:

[0458] The server adjusts the facial expression and pose of the generated avatar based on the emotional information obtained from the emotion engine. For example, if the user's emotion is recognized as "fun," the server adjusts the avatar's facial expression to a smile. The input data is the emotional information "fun," and the output data is an avatar image with a smiling face.

[0459] Step 6:

[0460] The server sends the final adjusted avatar image to the user's device, where the user can view the generated avatar and make further adjustments as needed. The input data is the adjusted avatar image, and the output data is the avatar image displayed on the user's device.

[0461] The above processing steps make it possible to generate profile pictures and avatars based on the user's characteristics and preferences, and also to perform matching that takes the user's emotions into consideration, thereby improving user satisfaction and matching accuracy.

[0462] Example 2

[0463] 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."

[0464] Conventional systems were unable to fully utilize the user's behavioral history and profile information, making it difficult to recommend optimal matches. Furthermore, because the system did not re-recommend matches based on the user's emotional changes, it was unable to increase user satisfaction. This made it difficult for users to continue using the system, and the success rate of matches also decreased.

[0465] 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.

[0466] In this invention, the server includes means for collecting user behavior history and profile information, means for preprocessing the collected data, means for inputting the preprocessed data into an AI model and recommending optimal matching partners, means for analyzing text entered by the user and tracking changes in emotions, and means for re-recommending matching partners based on changes in emotions. This makes it possible to recommend optimal matching partners based on the user's behavior history and profile information, and further, re-recommendations can be made taking into account changes in the user's emotions, thereby increasing user satisfaction and improving the success rate of matching. The server may also include means for tracking changes in the user's emotions.

[0467] "User behavior history" is a record of a series of operations and activities performed by a user within the system.

[0468] "Profile information" refers to information such as personal characteristics, preferences, and hobbies provided by a user.

[0469] "Preprocessing" is a series of operations that convert collected data into a format that is easy to analyze.

[0470] An "AI model" is an algorithm or system that uses artificial intelligence to analyze data and perform specific tasks.

[0471] A "matching partner" is another user who is recommended as a suitable partner for the user.

[0472] "Text analysis" is the process of analyzing the text entered by the user and understanding its content and sentiment.

[0473] "Changes in emotion" refers to fluctuations in emotion that can be read from the text entered by the user.

[0474] "Re-recommendation" refers to recommending a match partner again based on new information or conditions after the initial recommendation.

[0475] MODE FOR CARRYING OUT THE INVENTION

[0476] This invention is a system that collects user behavior history and profile information, and uses an AI model to recommend optimal matches based on that information. It also includes a function to track changes in the user's emotions and re-recommend matches based on those changes.

[0477] 1. Generate the system program

[0478] The server collects the user's behavioral history and profile information and stores it in a database. This includes information about the user's hobbies and preferences, and the people with whom they frequently communicate. The server inputs this data into an AI model and recommends the user the best possible match.

[0479] 2. Explain the generated program's processing in natural language

[0480] The server uses Python and machine learning libraries such as TensorFlow and PyTorch to analyze user data. For example, if a user's hobby is "watching movies" and they frequently exchange messages with other users who share the same hobby, the AI ​​will learn this tendency and recommend users who also share the hobby as potential matches.

[0481] The emotion engine also analyzes text entered by the user and tracks changes in emotion. The server uses a natural language understanding API as the emotion engine. For example, if a user enters "I'm happy today" and then enters "I'm lonely today" a short time later, the emotion engine recognizes the change and re-recommends matches that match the user's new emotion. The server may include means for tracking changes in the user's emotion. The server may also include means for re-recommending matches for the user based on the changes in the user's emotion.

[0482] 3. Add examples to your description

[0483] As a concrete example, assume that User A has hobbies of "watching movies" and "reading" and frequently communicates with other users who share these hobbies. The server collects User A's behavioral history and profile information and inputs it into an AI model. The AI ​​then recommends User B, whose hobbies are "watching movies" and "reading," as a potential match for User A.

[0484] For example, if User A inputs "I'm having fun today," and then a short time later inputs "I feel lonely today," the emotion engine will recognize this change. The server will then re-recommend User C, who also feels lonely, as a match that matches User A's new emotion.

[0485] Examples of prompts to input to a generative AI model might include:

[0486] Please recommend the best match based on the behavioral history and profile information of User A. User A's hobbies are "watching movies" and "reading," and he frequently communicates with other users who share these hobbies.

[0487] In this way, the processing of the system is explained while clarifying how the server, terminal, and user are involved.

[0488] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0489] Step 1:

[0490] User Data Collection

[0491] The server collects profile information and behavioral history provided by the user. Specifically, it stores the hobbies and preferences entered by the user, message exchanges, login history, etc. in a database. Input is performed by the user via a terminal, and output is user data stored on the server.

[0492] Step 2:

[0493] Data Preprocessing

[0494] The server preprocesses the collected data. Specifically, it performs tasks such as imputing missing values, normalizing data, and encoding categorical data. For example, it converts a user's hobbies, "watching movies" and "reading books," into numerical data. The input is the collected raw data, and the output is the preprocessed data.

[0495] Step 3:

[0496] Analysis and matchmaking recommendations using AI models

[0497] The server inputs the preprocessed data into the AI ​​model. The AI ​​model is built using TensorFlow and PyTorch. The model recommends optimal matches based on the user's behavioral patterns and profile information. For example, if user A's hobby is "watching movies," the model recommends user B who has the same hobby. The input is the preprocessed data, and the output is a list of recommended matches.

[0498] Step 4:

[0499] Emotion Tracking with Emotion Engine

[0500] The server sends the text entered by the user to the emotion engine. The emotion engine uses a natural language understanding API to analyze the user's emotion. For example, if the user enters "I'm having fun today," the emotion engine recognizes the emotion as "fun." The input is the text entered by the user, and the output is the analyzed emotion data.

[0501] Step 5:

[0502] Emotion-based re-recommendation

[0503] The server re-recommends matching partners based on the emotional data obtained from the emotion engine. For example, if a user inputs "I'm having fun today" and then "I'm lonely today," the server recognizes the change in emotion and re-recommends matching partners that fit the new emotion. The input is the emotional data from the emotion engine, and the output is a list of re-recommended matching partners.

[0504] (Application example 2)

[0505] 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."

[0506] Conventional systems match users based on their behavioral history and profile information, but do not consider changes in the user's emotions when matching or recommending content. As a result, they are unable to provide optimal matches or content that match the user's emotions, making it difficult to improve user satisfaction. Furthermore, there are also insufficient means to reduce bias in the amount of matching.

[0507] 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.

[0508] In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches using advanced attribute judgments using AI, means for reducing bias in the amount of matches, means for tracking changes in the user's emotions and re-recommending matching partners based on those changes, and means for analyzing the user's behavioral history and profile information and recommending optimal content. This makes it possible to provide optimal matching partners and content that match the user's emotions. The server may also include means for re-recommending matching partners for the user based on changes in the user's emotions.

[0509] "Means for creating profile pictures using AI" refers to a means for generating profile pictures using artificial intelligence based on the user's characteristics and preferences.

[0510] "Advanced attribute judgment using AI" is a method of analyzing a user's behavioral history and profile information and using artificial intelligence to make judgments to recommend the most suitable match.

[0511] "Means for inducing potential matches" refers to means for finding potential matches based on the user's attributes and behavioral history and recommending them to the user.

[0512] "Means for reducing bias in the amount of matching" refers to means for reducing bias in matching for specific users and achieving a balance in overall matching.

[0513] The "means for tracking changes in a user's emotions" refers to means for detecting changes in emotions from text entered by the user or actions of the user and recording the changes. The server may include means for re-recommending matches for the user based on the changes in the user's emotions.

[0514] The "means for re-recommending matching partners based on emotions" refers to a means for re-recommending appropriate matching partners based on changes in the user's emotions.

[0515] "Means for analyzing a user's behavior history and profile information" refers to means for collecting a user's past behavior and profile information and conducting analysis based on that information.

[0516] The "means for recommending optimal content" is a means for recommending optimal content to a user based on the user's behavioral history and changes in emotions.

[0517] A system for implementing this invention includes a server, a user terminal, and an AI model. The server collects user behavior history and profile information and performs advanced attribute judgments based on this data. Furthermore, the server tracks changes in the user's emotions and re-recommends matching partners and content based on those changes.

[0518] 1. System Program

[0519] The server runs a program that includes the following main means:

[0520] How profile pictures are created: Profile pictures are generated using AI based on user characteristics and preferences.

[0521] Advanced attribute determination methods: Analyzes user behavior history and profile information to recommend optimal matches.

[0522] Measures to reduce bias in matching volume: Reduce bias in matching towards specific users.

[0523] A means of tracking emotional changes: Detecting emotional changes from the text entered by the user and their actions, and recording those changes.

[0524] A method for re-recommending matching partners based on emotions: Based on changes in the user's emotions, appropriate matching partners are re-recommended.

[0525] A means of recommending optimal content: Recommending optimal content to users based on their behavioral history and emotional changes.

[0526] 2. Program processing explanation

[0527] The server implements these methods using a programming language such as Python. Specifically, it performs the following processes:

[0528] Profile picture creation: Using an AI model (e.g., GAN) to generate a profile picture based on the user's characteristics and preferences as input.

[0529] Advanced attribute determination: User behavioral history and profile information are retrieved from a database and analyzed using an AI model (e.g., TensorFlow or PyTorch).

[0530] Reduce bias in match volume: Adjust the matching algorithm to reduce bias in matches against certain users.

[0531] Sentiment change tracking: Analyzes user-entered text using natural language processing (NLP) techniques (e.g., BERT) to detect emotional changes.

[0532] Emotion-based re-recommendation: Based on changes in emotions, the AI ​​model is used again to recommend suitable matches.

[0533] Optimal content recommendation: Recommend optimal content using an AI model based on the user's behavioral history and emotional changes.

[0534] 3. Examples and prompts

[0535] As a concrete example, consider a scenario in which an application recommends a "Comedy Show" after a user inputs the emotion "fun." In this case, the server inputs the following prompt sentence into the generative AI model:

[0536] Example prompt sentence:

[0537] The user entered the emotion "fun." Please recommend content that matches their current emotion.

[0538] Using this prompt, the AI ​​model can recommend the most appropriate content based on the user's emotions.

[0539] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0540] Step 1:

[0541] The server collects user behavior history and profile information. It retrieves data entered by the user when using the application and past behavior history from the database. This provides input data for understanding the user's characteristics and preferences.

[0542] Step 2:

[0543] The server uses AI models to make sophisticated attribute judgments based on the collected user behavior history and profile information. Specifically, it uses machine learning frameworks such as TensorFlow and PyTorch to analyze the user's characteristics and recommend the most suitable matchmaking partners. This process outputs a list of the most suitable matches for the user.

[0544] Step 3:

[0545] The server tracks changes in the user's emotions. It analyzes the text data entered by the user using natural language processing (NLP) techniques (e.g., BERT) to detect changes in emotions. The input data is the text entered by the user in the application, and the output is the detected changes in emotions.

[0546] Step 4:

[0547] The server then uses the AI ​​model again to recommend suitable matches based on the changes in emotions. Using the changes in emotions as input data, the matching algorithm is run again to output matches that match the user's current emotions.

[0548] Step 5:

[0549] The server recommends optimal content based on the user's behavioral history and emotional changes. Using the user's behavioral history and emotional changes as input data, the server uses an AI model to select optimal content from a content database and recommend it to the user. This process outputs a list of content that is optimal for the user.

[0550] Step 6:

[0551] The user terminal displays a list of optimal matches and content received from the server. The user can select from the displayed matches and content and take the next action. This improves user satisfaction.

[0552] Example 3

[0553] 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."

[0554] In conventional systems, there was a tendency for bias in the amount of matching between users, which made it difficult for certain users to find a match. Furthermore, since the system did not take into account the user's emotions or mood when recommending a match, there was a possibility that user satisfaction would decrease. There is a need to solve these issues and provide a fairer and more satisfying matching experience.

[0555] 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.

[0556] In this invention, the server includes means for generating profile images using AI, means for inducing potential matches through advanced attribute analysis using AI, means for reducing bias in the amount of matches, means for collecting user emotion data and analyzing it with an emotion engine, and means for recommending matches based on the analysis results of the emotion engine. This reduces bias in the amount of matches for users and enables recommendation of matches based on emotions.

[0557] "AI-based profile image generation means" is a function that uses artificial intelligence to automatically generate profile images based on the user's characteristics and preferences.

[0558] "Advanced attribute analysis using AI" is the process of using artificial intelligence to analyze a user's behavioral history and profile information to identify the most suitable match.

[0559] "Means to induce potential matches" is a function in which artificial intelligence recommends potential compatible match partners based on the user's attributes and behavioral data.

[0560] The "means for reducing the imbalance in the number of matches" is a function for monitoring the matching status of all users and actively recommending potential matches to users with fewer matches. The server may include means for monitoring the matching status of all users and recommending potential matches to users with fewer matches than a specific user.

[0561] The "means for collecting user emotional data" is a function for collecting text data and behavioral data entered by the user and understanding the user's emotional state.

[0562] "Means for analyzing with an emotion engine" refers to the process of analyzing collected user emotion data using an emotion engine and learning changes in the user's emotions.

[0563] The "means for recommending a match partner based on the analysis results of the emotion engine" is a function that recommends a match partner that matches the user's emotional state based on the analysis results of the emotion engine.

[0564] MODE FOR CARRYING OUT THE INVENTION

[0565] This invention is a system that uses AI to generate profile images, performs advanced attribute analysis, reduces bias in the amount of matches, and recommends matching partners based on emotions using an emotion engine, in order to improve users' matching experience.

[0566] Hardware and software used

[0567] The server uses the following hardware and software:

[0568] Hardware: High-performance server machine (CPU, GPU, memory, storage)

[0569] Software: TensorFlow, PyTorch, database management system (DBMS), mail server, push notification service, Google® Natural Language API, IBM Watson® sentiment analysis tool

[0570] Terminals are devices used by users (smartphones, tablets, PCs) that use the following software:

[0571] Software: Chat app, diary app, data collection module

[0572] Program processing

[0573] The server periodically collects the matching history data of all users and identifies users who have not had a match for a certain period of time (for example, one month). For these identified users, AI is used to recommend potential matches. The AI ​​model used here is built using TensorFlow and PyTorch.

[0574] Examples:

[0575] If user A has no matches for a month, the server will recommend users B and C based on user A's attributes (age, hobbies, region, etc.).

[0576] The device collects text data entered by the user, for example, by saving text entered through a chat app or diary app, and periodically transmits the text to a server.

[0577] The server analyzes the user's text data and learns changes in emotions using an emotion engine, which uses Google's Natural Language API and IBM Watson's emotion analysis tools.

[0578] Examples:

[0579] If user D tends to input the text "fun" every weekend, the server learns this tendency and recommends user E, who matches the "fun" mood, to user D on the weekend.

[0580] Prompt Sentence Examples

[0581] Here are some example prompts to input to the generative AI model:

[0582] "For users who have not had a match for a month, please recommend users who are likely to match their attributes."

[0583] "If a user tends to type 'fun' on the weekends, recommend matches that fit that user."

[0584] In this way, the system can improve the matching experience of the user. The flow of the identification process in the third embodiment will be described with reference to FIG.

[0585] Program processing flow

[0586] Step 1: Monitor your users' matches

[0587] The server periodically collects the matching history data of all users. As input, it retrieves the matching history of each user from the database. As output, it generates a list of users who have not had a match for a certain period of time (e.g., one month).

[0588] Specific behavior:

[0589] The server runs a script every night at midnight to check the matching history of all users.

[0590] The server lists users who have not had a match for more than a month.

[0591] Step 2: Make recommendations for users with few matches

[0592] The server uses AI to recommend potential matches for the identified user. The input is the user's attribute data (age, hobbies, region, etc.) and is input into the AI ​​model. The output is a list of recommended users.

[0593] Specific behavior:

[0594] The server inputs the identified user's attribute data into the AI ​​model.

[0595] The server obtains the recommended user list output from the AI ​​model and stores it in a database.

[0596] Step 3: Notify the user of the recommendation

[0597] The server notifies the user of the recommendation results. As input, it generates a notification message based on the recommended user list. As output, it sends a notification to the user via email or in-app notification.

[0598] Specific behavior:

[0599] The server generates a notification message based on the recommended user list.

[0600] The server sends notifications to the user through a mail server or push notification service.

[0601] Step 4: Collect user text data

[0602] The device collects text data entered by the user. As input, it stores the text entered through a chat app or diary app. As output, it sends the collected text data to a server.

[0603] Specific behavior:

[0604] The device captures text entered by the user in real time and stores it in local storage.

[0605] The terminal periodically transmits text data to the server.

[0606] Step 5: Learn emotional changes with the emotion engine

[0607] The server analyzes the user's text data using an emotion engine and learns changes in emotions. The received text data is input to the emotion engine. As an output, the emotion analysis results are generated and stored in a database.

[0608] Specific behavior:

[0609] The server inputs the received text data into the emotion engine.

[0610] The server stores the emotion analysis results output from the emotion engine in a database.

[0611] Step 6: Recommend matches based on emotions

[0612] The server recommends a match that matches the user's emotions based on the learning results of the emotion engine. As input, it identifies the user's emotional patterns based on the emotion analysis results. As output, it uses an AI model to recommend other users with similar emotional patterns and generates a recommendation list.

[0613] Specific behavior:

[0614] The server identifies the user's emotional pattern based on the emotion analysis results.

[0615] The server uses an AI model to recommend other users with similar emotional patterns and generate a recommendation list.

[0616] The server notifies the user of the recommendation list.

[0617] (Application example 3)

[0618] 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."

[0619] In conventional systems, there is often a bias in the amount of matching between users, making it difficult for certain users to find a match. In addition, because recommendations do not utilize individual data such as user emotions or order history, it is difficult to provide the optimal match or recommendation for the user. This leads to issues such as lower user satisfaction and a low rate of continued use of the system.

[0620] 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.

[0621] In this invention, the server includes means for creating profile pictures using AI, means for inducing potential matches using advanced attribute judgments by AI, means for reducing bias in the amount of matches, means for making optimal recommendations based on the user's order history and emotional data, and means for the emotion engine to learn changes in the user's emotions and make recommendations based on the learning results. This enables optimal matching and recommendations using individual user data, improving user satisfaction and increasing the rate at which the system is used.

[0622] "AI-based profile picture creation method" is a function that uses artificial intelligence to generate profile pictures based on the user's characteristics and preferences.

[0623] "Advanced attribute judgment using AI" is a function that analyzes a user's behavioral history and profile information and uses artificial intelligence to recommend the most suitable match.

[0624] "Means to reduce imbalances in the amount of matches" refers to a function that monitors the matching status of all users and actively recommends potential matches to users with few matches.

[0625] "Means for making optimal recommendations based on a user's order history and emotional data" is a function that analyzes a user's past order history and emotional data and recommends optimal restaurants and menus based on that.

[0626] "A means for the emotion engine to learn changes in the user's emotions and make recommendations based on the learning results" is a function that learns the user's emotional data and makes optimal matches and recommendations based on those changes.

[0627] To implement the present invention, the following system configuration and program are required.

[0628] System Configuration

[0629] 1. Server: The server generates a user profile picture and runs an AI model to analyze the user's behavioral history and emotional data. The server accesses the database to obtain the user's order history and emotional data, and then makes optimal recommendations based on that.

[0630] 2. User Device: User devices include smartphones, smart glasses, head-mounted displays, or robots. These devices receive the recommendation results from the server and display them to the user.

[0631] Program processing explanation

[0632] The server performs the following process.

[0633] 1. Data collection and storage: The server stores user profile information, behavioral history, order history, and emotion data in a database, including text data entered by users and order history data.

[0634] 2. Profile photo generation: The server generates a profile photo based on the user's characteristics and preferences using an AI model. This AI model generates the optimal photo based on the user's input data.

[0635] 3. Advanced attribute judgment: The server analyzes the user's behavioral history and profile information and recommends the best match using an AI model that identifies potential matches based on the user's past behavioral data.

[0636] 4. Mitigating the imbalance in the amount of matches: The server monitors the matching status of all users and proactively recommends potential matches to users with few matches, thereby reducing the imbalance in the amount of matches.

[0637] 5. Emotional Data Learning and Recommendation: The server uses an emotion engine to learn changes in the user's emotions and makes optimal recommendations based on the learning results. For example, if a user tends to feel "happy" on weekends, the server will recommend restaurants and menus that match the user's happy mood on weekends.

[0638] Specific examples

[0639] As a specific example, consider the case of recommending a restaurant to a user with a user ID of 1. Assume that this user tends to have the emotion "fun" on weekends.

[0640] Example prompt sentence:

[0641] For user ID 1, if the user tends to have the emotion "fun" on weekends, recommend the best restaurant.

[0642] By inputting this prompt into a generative AI model, restaurant recommendations are made based on the user's emotional data.

[0643] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0644] Step 1:

[0645] Data collection and storage

[0646] The server collects user profile information, behavioral history, order history, and emotional data and stores them in a database. Specifically, it acquires text data entered by the user into the application and order history data, and stores this data in the database. The input is the user's profile information, order history, and emotional data, and the output is this data stored in the database.

[0647] Step 2:

[0648] Generate a profile photo

[0649] The server generates a profile picture using an AI model based on the user's characteristics and preferences. Specifically, the server inputs the user's input data (e.g., favorite colors and styles) into the AI ​​model to generate an optimal profile picture. The input is the user's characteristics and preferences, and the output is the generated profile picture.

[0650] Step 3:

[0651] Advanced attribute judgment

[0652] The server analyzes the user's behavioral history and profile information and uses an AI model to recommend optimal matches. Specifically, the user's past behavioral data is input into the AI ​​model to identify potential matches. The input is the user's behavioral history and profile information, and the output is recommended matches.

[0653] Step 4:

[0654] Reducing bias in matching volume

[0655] The server monitors the matching status of all users and proactively recommends new matches to users with few matches. Specifically, it analyzes the matching data of all users, identifies users with few matches, and recommends new matches to those users. The input is the matching data of all users, and the output is the recommended matches.

[0656] Step 5:

[0657] Emotional Data Learning and Recommendation

[0658] The server uses an emotion engine to learn changes in the user's emotions and makes optimal recommendations based on the learning results. Specifically, the user's emotional data is input into the emotion engine, which learns changes in emotions. Then, based on the learning results, the server recommends the optimal restaurant and menu for the user. The input is the user's emotional data, and the output is the recommended restaurant and menu.

[0659] Step 6:

[0660] Displaying recommendation results

[0661] The user terminal receives the recommendation results from the server and displays them to the user. Specifically, the recommendation results sent from the server are received by the user terminal and displayed to the user through the application. The input is the recommendation results from the server, and the output is the recommendation results displayed to the user.

[0662] 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.

[0663] 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.

[0664] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are mentioned.

[0665] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing is performed by the smart device 14.

[0666] This may be done.

[0667] [Second embodiment]

[0668] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0669] 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.

[0670] 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).

[0671] 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.

[0672] 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.

[0673] 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).

[0674] 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.

[0675] 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.

[0676] 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.

[0677] 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.

[0678] 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.

[0679] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0680] "Example 1"

[0681] The system of the present invention utilizes AI technology to create a profile picture for a user. Specifically, the AI ​​generates the profile picture based on information such as a self-introduction, hobbies, and preferences provided by the user. For example, if a user provides information that they "like the outdoors," the AI ​​will set the background of the user's profile picture to an outdoor scene, such as a campsite or mountain.

[0682] "Example 2"

[0683] The system of the present invention also uses AI to perform advanced attribute judgments and guide potential matches. Specifically, the AI ​​analyzes the user's behavioral history and profile information to recommend optimal matches. For example, if a user tends to frequently communicate with people who share specific hobbies or preferences, the AI ​​can learn that tendency and recommend people with similar hobbies and preferences as potential matches.

[0684] "Example 3"

[0685] Furthermore, the system of the present invention reduces the bias in the amount of matches. Specifically, the AI ​​monitors the matching status of all users and more proactively recommends potential matches to users with few matches. For example, if a user has not had a match for a certain period of time, the AI ​​will proactively recommend users who are likely to match the user's attributes.

[0686] The processing flow of each embodiment will be described below.

[0687] "Example 1"

[0688] Step 1: The user provides information such as a self-introduction, hobbies, and preferences.

[0689] Step 2: The AI ​​generates a profile picture based on the provided information. For example, if a user provides information that they "like the outdoors," the AI ​​will set the user's profile picture with an outdoor scene, such as a campsite or mountain, as the background.

[0690] Step 3: The generated profile picture is set as the user's profile.

[0691] Step 1: The user's behavioral history and profile information are analyzed by AI.

[0692] Step 2: The AI ​​recommends optimal matches based on the analysis results. For example, if a user tends to frequently communicate with people who share certain hobbies or preferences, the AI ​​will learn that tendency and recommend people with similar hobbies and preferences as potential matches.

[0693] "Example 3"

[0694] Step 1: AI monitors the matching status of all users.

[0695] Step 2: For users with few matches, the AI ​​will more proactively recommend potential matches. For example, if a user has not had a match for a certain period of time, the AI ​​will proactively recommend users who are likely to match the user's attributes.

[0696] Example 1

[0697] 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."

[0698] In conventional systems, users' profile pictures were based on self-reporting, and often did not fully reflect the user's characteristics and preferences. Furthermore, it was difficult to consider the user's latent attributes and behavioral history when matching, resulting in low matching accuracy. Furthermore, there was also the issue of matching being biased toward certain users, resulting in an imbalance in the overall amount of matches.

[0699] 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.

[0700] In this invention, the server includes means for receiving profile information entered by a user, means for generating a prompt sentence based on the received profile information, means for inputting the generated prompt sentence into a generative AI model and generating a profile picture, means for sending the generated profile picture to the user's device, means for guiding potential matches through advanced attribute judgment by AI, and means for reducing bias in the amount of matches. This makes it possible to generate profile pictures that reflect the user's characteristics and preferences, improving matching accuracy and balancing the overall amount of matches.

[0701] The "means for receiving profile information entered by the user" is a function that allows the server to receive information such as a self-introduction, hobbies, and preferences entered by the user through the terminal.

[0702] "Means for generating a prompt sentence based on received profile information" refers to a function that creates a prompt sentence to be input into the generative AI model based on the user's profile information received by the server.

[0703] "Means for inputting the generated prompt sentence into a generative AI model and generating a profile picture" refers to a function that inputs the prompt sentence generated by the server into a generative AI model, thereby generating a profile picture that matches the user's characteristics and preferences.

[0704] "Means for sending the generated profile picture to the user's terminal" is a function that enables the server to send the profile picture obtained from the generative AI model to the user's terminal.

[0705] "Means of guiding potential matches through advanced attribute judgment by AI" is a function in which AI analyzes a user's behavioral history and profile information, determines the user's potential attributes, and recommends the most suitable match.

[0706] "Means for reducing bias in the amount of matching" is a function for balancing the overall amount of matching so that matching is not biased towards specific users.

[0707] MODE FOR CARRYING OUT THE INVENTION

[0708] The present invention provides a system for generating a profile picture based on a user's characteristics and preferences, thereby improving the accuracy of matching. Specific embodiments of this system will be described below.

[0709] Program Generation

[0710] The server generates a program to generate a profile picture for the user. This program uses a generative AI model to create the profile picture based on the information provided by the user. Examples of generative AI models that can be used include OpenAI's DALL-E and MidJourney.

[0711] Program processing

[0712] The server receives profile information entered by the user through the terminal. This profile information includes a self-introduction, hobbies, preferences, etc. For example, if a user enters "I like the outdoors," this information is sent to the server.

[0713] Next, the server generates a prompt based on the received profile information. The prompt is text to be input into the generative AI model, and reflects the user's characteristics and preferences. For example, a prompt might be generated that reads, "Please generate a profile picture for a user who loves the outdoors. Please set the background to an outdoor scene, such as a campsite or mountain."

[0714] The server inputs the generated prompt sentence into the generative AI model, which then generates a profile picture that matches the user's characteristics and preferences based on the prompt sentence. The generated profile picture is then sent to the server.

[0715] The server sends the generated profile picture to the user's device, and the user can check the received profile picture on the device and use it as their profile.

[0716] Specific examples

[0717] For example, if a user enters the information "I like reading," the following prompt will be generated:

[0718] "Generate a profile picture for a user who likes reading. Set the background to a reading scene, such as a library or bookshelf."

[0719] Based on this prompt, the generative AI model generates a profile picture with a library and bookshelves as the background, and the generated picture is sent from the server to the user's device, where the user can use the picture as their profile picture.

[0720] In this way, by generating a profile picture based on the user's characteristics and preferences, the accuracy of matching can be improved. Furthermore, advanced attribute judgment using AI can analyze the user's behavioral history and profile information to recommend the most suitable match. Furthermore, by including a means to reduce bias in the amount of matches, the overall amount of matches can be balanced.

[0721] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0722] Step 1:

[0723] The user enters profile information.

[0724] The user uses the terminal to input profile information such as a self-introduction, hobbies, preferences, etc. For example, the user may input "I like the outdoors." This input information is sent to the server.

[0725] Step 2:

[0726] The server receives the user's input information.

[0727] The server receives the profile information sent by the user from the terminal. This information is text data that reflects the user's hobbies and preferences. The received data is used in the next processing step.

[0728] Step 3:

[0729] The server generates a prompt sentence based on the received profile information.

[0730] The server creates a prompt to input to the generative AI model based on the received user profile information. For example, it generates a prompt such as, "Please generate a profile picture for a user who likes the outdoors. Please set the background to an outdoor scene, such as a campsite or mountain." This prompt is used in the next step.

[0731] Step 4:

[0732] The prompt sentence generated by the server is input into the generative AI model.

[0733] The server inputs the generated prompt into a generative AI model (such as OpenAI's DALL-E or MidJourney), which then generates a profile picture that matches the user's characteristics and preferences based on the prompt. The generative AI model then analyzes the prompt and generates an image with an appropriate background and scene.

[0734] Step 5:

[0735] The server receives the generated profile photo.

[0736] The server receives a profile picture generated from the generative AI model, which includes a background and scene that matches the user's interests and preferences. The received profile picture is used in the next step.

[0737] Step 6:

[0738] The server sends the profile picture to the user's device.

[0739] The server sends the received profile picture to the user's device, where the user can use the picture as their profile. The sent picture is displayed on the user's device.

[0740] Step 7:

[0741] The user checks their profile picture.

[0742] The user can review the received profile photo on their device, check if the photo meets their expectations, and request a regeneration if necessary.

[0743] (Application example 1)

[0744] 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."

[0745] In conventional systems, users' profile pictures are based on self-reporting and do not reflect their hobbies or preferences, which means that they cannot fully express the user's personality and interests. Another issue is that poor quality profile pictures result in a lower matching success rate. Furthermore, because the system does not fully utilize the user's behavioral history and profile information, it is difficult to recommend optimal matches.

[0746] 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.

[0747] In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches through advanced attribute judgment by AI, means for reducing bias in the amount of matches, means for generating prompt sentences based on the user's hobbies and preferences and generating a profile picture using a generative AI model, and means for reflecting the generated profile picture in the user's profile. This makes it possible to generate a high-quality profile picture that reflects the user's personality and interests and recommend optimal matches.

[0748] An "AI profile picture generation method" is a method that uses artificial intelligence technology to generate a profile picture based on user-provided characteristics and preferences.

[0749] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match.

[0750] "Means for inducing potential matches" are means for finding potentially compatible partners based on the user's attributes and behavioral history, and promoting matching.

[0751] "Means for reducing bias in the amount of matching" refers to means for evenly distributing the overall amount of matching so that matching is not concentrated on specific users.

[0752] "Means for generating prompt sentences based on the user's hobbies and preferences" refers to means for creating prompt sentences to be input into a generative AI model based on information about hobbies and preferences provided by the user.

[0753] "Means for generating a profile picture using a generative AI model" refers to means for generating a profile picture for a user using a generative AI model based on a generated prompt sentence.

[0754] The "means for reflecting the generated profile picture in the user's profile" refers to a means for adding the generated profile picture to the user's profile information and displaying it.

[0755] The system for implementing this invention is composed of three main elements: a server, a terminal, and a user. The specific operations of each element and their cooperation will be explained below.

[0756] server

[0757] The server includes a means for creating a profile picture using AI, a means for determining advanced attributes using AI, a means for guiding potential matches, a means for reducing bias in the amount of matches, a means for generating prompt sentences based on the user's hobbies and preferences, a means for generating a profile picture using a generative AI model, and a means for reflecting the generated profile picture in the user's profile.

[0758] The server receives the user's hobbies and preferences and generates a prompt based on them. Using the generated prompt, it uses a generative AI model (e.g., OpenAI's API) to generate a profile picture for the user. The generated profile picture is added to the user's profile information and displayed.

[0759] Terminal

[0760] The terminal is a device where the user enters information and checks the generated profile picture. It can be a smartphone or a PC. The user enters information about their hobbies and preferences through the terminal and sends it to the server. The generated profile picture can be checked on the terminal.

[0761] User

[0762] A user is an individual who uses the system. The user inputs information about their hobbies and preferences through their device and sends it to the server. The server generates a profile picture, which the user can then view on their device.

[0763] Specific examples

[0764] For example, if the user provides information such as "I like fashion," the server generates the following prompt sentence:

[0765] Prompt Sentence Examples

[0766] The user's hobby is fashion, and their preference is stylish cityscapes.

[0767] Based on this prompt, a generative AI model is used to generate a profile picture featuring a fashion show or stylish cityscape as a background, which is then added to the user's profile information and can be viewed on the device.

[0768] In this way, it is possible to generate high-quality profile pictures that reflect the user's personality and interests, and to recommend optimal matches.

[0769] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0770] Step 1:

[0771] The user uses the terminal to input information about his or her hobbies and preferences.

[0772] Input: User's hobbies and preferences (e.g., "I like fashion")

[0773] Output: The entered information is sent to the server.

[0774] Specific operation: The user enters information about their hobbies and preferences into the input form on the device and presses the "Send" button. The device then sends this information to the server.

[0775] Step 2:

[0776] The server generates a prompt based on the received user information.

[0777] Input: User's hobbies and preferences

[0778] Output: Generated prompt sentence (e.g., "The user's hobby is fashion, and their preference is stylish cityscapes.")

[0779] Specific operation: The server analyzes the received information and generates a prompt sentence based on the template.

[0780] Step 3:

[0781] The server sends a request to the generative AI model using the generated prompt sentence.

[0782] Input: Generated prompt text

[0783] Output: Profile photo URL returned by the generative AI model

[0784] Specific operation: The server sends the generated prompt to a generative AI model (e.g., OpenAI's API) and requests the generation of a profile photo. The generative AI model generates an image based on the prompt and returns its URL to the server.

[0785] Step 4:

[0786] The server receives the profile picture URL returned by the generative AI model and downloads the image.

[0787] Input: Profile photo URL

[0788] Output: Downloaded profile photo

[0789] What happens: The server uses the returned URL to download the profile picture.

[0790] Step 5:

[0791] The server adds the downloaded profile picture to the user's profile information and sends it to the terminal.

[0792] Input: Downloaded profile photo

[0793] Output: Updated user profile information

[0794] Specific operation: The server adds the downloaded profile picture to the user's profile information and sends the updated profile information to the terminal.

[0795] Step 6:

[0796] The user uses the device to view the generated profile picture.

[0797] Input: Updated user profile information

[0798] Output: The generated profile picture displayed on the device.

[0799] Specific operation: The user checks the generated profile picture through the device, and the device displays the updated profile information received from the server.

[0800] Example 2

[0801] 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."

[0802] Conventional systems have the problem of being unable to fully utilize users' behavioral history and profile information, resulting in low accuracy in recommending optimal matches. Furthermore, insufficient data preprocessing reduces the learning efficiency of the AI ​​model, resulting in poor quality recommendation results.

[0803] 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.

[0804] In this invention, the server includes means for collecting user behavior history and profile information, means for preprocessing the collected data, means for training an AI model using the preprocessed data, and means for recommending optimal matching partners using the trained AI model. This makes it possible to effectively utilize the user's behavior history and profile information and recommend optimal matching partners with high accuracy.

[0805] "User behavior history" is a record of the operations and activities that a user performs on the system.

[0806] "Profile information" refers to personal information and attribute information provided by a user at the time of registration.

[0807] "Means of collection" refers to the methods and technologies used to store user behavior history and profile information in a database.

[0808] "Preprocessing means" refers to methods and techniques that process collected data through cleaning, normalization, feature engineering, etc., to convert it into a format that is easy for an AI model to learn.

[0809] "Means for training an AI model" refers to methods or techniques that use preprocessed data to apply machine learning algorithms to learn user behavior patterns and attributes.

[0810] "Means for recommending optimal matching partners" refers to methods and technologies that use trained AI models to predict and present optimal matching partners to users.

[0811] This invention is a system that recommends optimal matches by collecting user behavior history and profile information and generating an AI model based on this information. Specific embodiments of this system are described below.

[0812] First, the server collects the user's behavioral history and profile information. The behavioral history includes the actions the user took on the website or application (e.g., page views, message sending, likes). The profile information includes the personal information and attribute information (e.g., age, gender, hobbies, occupation) entered by the user when registering. This data is stored on a database server (e.g., MySQL, PostgreSQL). A data collection tool (e.g., Apache Kafka) is used to collect the data.

[0813] Next, the server preprocesses the collected data. Preprocessing includes data cleaning (e.g., imputing missing values ​​and removing outliers), normalization (e.g., scaling numerical data), and feature engineering (e.g., generating new features). This converts the data into a format suitable for training AI models. Data preprocessing tools (e.g., Pandas and NumPy) are used for data preprocessing.

[0814] The server then uses the preprocessed data to train an AI model using a machine learning framework (e.g., TensorFlow, PyTorch). The AI ​​model learns the user's behavioral patterns and attributes to predict the best possible match.

[0815] Finally, the server uses the trained AI model to recommend the best possible match for the user. Specifically, the user's behavioral history and profile information are input into the AI ​​model, and the model presents the best possible match to the user. The recommendation system uses recommendation algorithms (e.g., Scikit-learn, LightGBM).

[0816] Examples:

[0817] If user A has hobbies of "hiking" and "cooking" and frequently communicates with other users who have these hobbies, the AI ​​will learn this tendency. Next, if user B also has hobbies of "hiking" and "cooking," the AI ​​will recommend user A and user B as potential matches.

[0818] Example prompt sentence:

[0819] "Generate an AI model that recommends optimal matches based on a user's behavioral history and profile information. User A's hobbies are hiking and cooking, and he frequently communicates with other users who share these hobbies. User B's hobbies are also hiking and cooking. Based on this information, create an AI model that recommends User A and User B as potential matches."

[0820] The above is a specific embodiment for carrying out the present invention.

[0821] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0822] Step 1:

[0823] Data collection

[0824] The server collects the user's behavioral history and profile information.

[0825] Input: User operation history (e.g., page viewing, message sending, likes) and registration information (e.g., age, gender, hobbies, occupation)

[0826] Specific operations: The server logs the actions users take on websites and applications and stores them in a database. Data collection tools (e.g., Apache Kafka) are used to collect data in real time.

[0827] Output: User behavior history and profile information stored in a database

[0828] Step 2:

[0829] Data Preprocessing

[0830] The server pre-processes the collected data.

[0831] Input: User behavior and profile information stored in a database

[0832] Specific operations: The server performs data cleaning (e.g., missing value imputation, outlier removal), normalization (e.g., scaling numerical data), and feature engineering (e.g., generating new features) using data preprocessing tools (e.g., Pandas, NumPy).

[0833] Output: Preprocessed dataset

[0834] Step 3:

[0835] Training an AI model

[0836] The server uses the preprocessed data to train the AI ​​model.

[0837] Input: Preprocessed dataset

[0838] Specific operation: The server uses machine learning algorithms (e.g., deep learning, gradient boosting) to learn user behavior patterns and attributes. It uses machine learning frameworks (e.g., TensorFlow, PyTorch).

[0839] Output: A trained AI model

[0840] Step 4:

[0841] Matching partner recommendations

[0842] The server uses a trained AI model to recommend the best possible match for the user.

[0843] Input: Trained AI model, user behavior history and profile information

[0844] Specific operation: The server inputs the user's behavioral history and profile information into the AI ​​model, and presents the user with the best match predicted by the model. It uses a recommendation system (e.g., Scikit-learn, LightGBM).

[0845] Output: A list of best matches presented to the user

[0846] The above is the specific processing flow of the program of this system.

[0847] (Application example 2)

[0848] 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."

[0849] Conventional systems were unable to fully utilize a user's behavioral history and profile information, making it difficult to recommend optimal matches and products. Furthermore, there was a tendency for matching volume to be biased, reducing user satisfaction. Furthermore, it was difficult to make highly accurate product recommendations based on a user's interests and preferences.

[0850] 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.

[0851] In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches using advanced attribute judgments using AI, means for reducing bias in the amount of matches, means for analyzing the user's behavioral history and profile information to recommend optimal products, and means for vectorizing product descriptions and calculating cosine similarity to recommend similar products. This makes it possible to recommend optimal matches and products with high accuracy based on the user's behavioral history and profile information.

[0852] "AI-based profile picture creation method" refers to a method that uses artificial intelligence technology to generate a profile picture based on a user's characteristics and preferences.

[0853] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match and products.

[0854] "Means for inducing potential matches" refers to means for finding and recommending potential matches based on the user's attributes and behavioral history.

[0855] "Means for reducing bias in the amount of matching" refers to means for reducing bias in matching for particular users and providing equal opportunities for matching.

[0856] "Means for analyzing a user's behavioral history and profile information and recommending optimal products" refers to means for recommending optimal products to a user based on the user's past behavior and profile information.

[0857] "Means for vectorizing product descriptions, calculating cosine similarity, and recommending similar products" refers to a means for finding and recommending similar products by converting product descriptions into numerical vectors and calculating cosine similarity.

[0858] A system for implementing this invention includes a server, a user terminal, and an AI model. The server collects user behavioral history and profile information and runs software to analyze this data. Specifically, the server processes the data using libraries such as Python, Pandas, and Scikit-learn.

[0859] The server first loads the user's behavioral history data and product data. Next, it uses TfidfVectorizer to vectorize the product descriptions and extract the user's interests. This allows it to obtain vectors of products that the user has shown interest in in the past. It then calculates the cosine similarity and recommends the most similar products.

[0860] The user terminal is a device such as a smartphone or PC that displays the recommendation results sent from the server. The user can then check and purchase the recommended products.

[0861] For example, if a user frequently purchases "sports goods," the server will recommend the latest sports goods and related products based on that user's behavioral history, allowing the user to easily find products that match their interests.

[0862] An example of a prompt is as follows:

[0863] "Create a Python program that recommends optimal products based on user behavior history data. Include a function to extract user interests and calculate cosine similarity to recommend the most similar products."

[0864] In this way, the system for implementing the invention is able to recommend optimal matches and products with a high degree of accuracy based on the user's behavioral history and profile information.

[0865] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0866] Step 1:

[0867] The server reads user behavior history data and product data. Specifically, it obtains data from files named user_behavior.csv and product_catalog.csv. The input is user behavior history data and product data, and the output is a data frame containing these data. The server reads these data using the Pandas library.

[0868] Step 2:

[0869] The server vectorizes the product description. Specifically, it converts the product description into a numeric vector using TfidfVectorizer. The input is the product description text, and the output is the vectorized product description. The server performs this process using the Scikit-learn library.

[0870] Step 3:

[0871] The server obtains vectors of products of interest from the user's behavioral history. Specifically, it extracts corresponding vectors based on the IDs of products that the user has previously shown interest in. The input is the user's behavioral history data and vectorized product descriptions, and the output is a vector that reflects the user's interests.

[0872] Step 4:

[0873] The server calculates the cosine similarity between the vector reflecting the user's interests and the vector of all products. The input is the vector reflecting the user's interests and the vector of all products, and the output is the similarity score. The server performs this calculation using the Scikit-learn library.

[0874] Step 5:

[0875] The server recommends the most similar products. Specifically, it selects the product with the highest score based on the cosine similarity score and recommends it to the user. The input is the similarity score, and the output is a list of recommended products.

[0876] Step 6:

[0877] The terminal displays the recommendation results sent from the server. Specifically, it displays a list of recommended products on the user's smartphone or PC. The input is the recommendation results sent from the server, and the output is the product list displayed to the user.

[0878] Step 7:

[0879] The user checks the recommended products and purchases them. Specifically, they select products of interest from the product list displayed on the terminal and then carry out the purchase procedure. The input is the list of recommended products, and the output is information about the purchased product.

[0880] Example 3

[0881] 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."

[0882] In conventional systems, there was a problem that the amount of matching between users was easily biased, and some users did not get matching opportunities. In addition, there was an issue that appropriate recommendations were not made to users with few matches, resulting in a poor user experience.

[0883] 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.

[0884] In this invention, the server includes means for storing the matching status of users in a database, means for the server to periodically acquire the matching status of all users, means for the server to input the matching status to the AI ​​model, means for the AI ​​model to identify users with few matches, means for the AI ​​model to select recommended users for the identified users, and means for the server to notify target users of information about the recommended users. This makes it possible to reduce bias in the amount of matches and make appropriate recommendations to users with few matches.

[0885] "Means for saving a user's matching status in a database" is a function that records information about a new match made by a user in a database.

[0886] "Means for the server to periodically obtain the matching status of all users" refers to a function by which the server obtains the matching status of all users from the database at regular time intervals.

[0887] "Means for the server to input the matching status into the AI ​​model" refers to the function by which the server inputs the matching status data it has acquired into the AI ​​model.

[0888] "Means for the AI ​​model to identify users with few matches" refers to a function in which the AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time.

[0889] "Means for selecting recommended users for a user identified by an AI model" is a function that selects recommended users for a user identified by an AI model based on attributes and past matching history.

[0890] "Means for the server to notify the target user of information about recommended users" is a function by which the server notifies the target user of information about recommended users selected by the AI ​​model.

[0891] The "scheduler" is a function that allows the server to automate the process of periodically obtaining the matching status of all users.

[0892] "User attributes" refers to personal information such as the user's age, hobbies, and region.

[0893] MODE FOR CARRYING OUT THE INVENTION

[0894] The present invention relates to a system that reduces bias in the amount of matching for users and makes appropriate recommendations to users with few matches. Specific embodiments of this system will be described below.

[0895] System configuration

[0896] This system consists of three main elements: a server, a terminal, and a user. The server contains software components such as a database, an AI model, and a scheduler. A terminal is a device used by a user (such as a smartphone, tablet, or PC). A user is an individual who uses the system.

[0897] Hardware and software used

[0898] Server: Use a high-performance server machine (e.g., AWS EC2 instance). Use MySQL as the database and TensorFlow or PyTorch to build the AI ​​model.

[0899] Device: A user device such as a smartphone, tablet, or PC.

[0900] Scheduler: Use scheduling tools such as cron jobs to automate periodic data retrieval.

[0901] Program processing

[0902] The server stores the matching status of users in a database. When a user makes a new match, the device records that information in the database. The server retrieves the matching status of all users from the database at regular intervals (e.g., every hour). This process is automated using a scheduler.

[0903] The server inputs the acquired matching status data into the AI ​​model. The AI ​​model performs analysis based on the user's attributes (age, hobbies, region, etc.) and past matching history. The AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time (e.g., one week). These identified users are listed as users with few matches.

[0904] The AI ​​model recommends other users who are likely to have matching attributes and hobbies to the identified user. The recommendation algorithm is based on the user's attributes, such as age, hobbies, and region. The server notifies the target user of information about the recommended users selected by the AI ​​model. Notifications are sent to the device via push notification or email. By receiving the notification, users gain new matching opportunities.

[0905] Specific examples

[0906] For example, if User A (age 30, hobby: hiking, location: Tokyo) has not had a match for more than a week, the server retrieves User A's matching status from the database. The server inputs this data into the AI ​​model, which then recommends User B (age 28, hobby: hiking, location: Tokyo), who is likely to be a good match for User A. The server then sends User A a push notification with User B's profile. User A receives the notification and has the opportunity to newly match with User B.

[0907] Prompt Sentence Examples

[0908] "If User A, age 30, has not had a match for more than a week, the AI ​​should recommend User B, age 28, who enjoys hiking."

[0909] In this way, the system processing can be performed while clarifying the roles of the server, terminal, and user. The flow of the specific processing in the third embodiment will be described with reference to FIG.

[0910] Step 1:

[0911] When a user makes a new match, the device stores that information in a database.

[0912] Input: User's matching information (user ID, matched person's ID, matching date and time)

[0913] Output: Matching information stored in a database

[0914] Specific operation: When a user matches, the device records the user ID, the matched person's ID, and the date and time of the match in a database.

[0915] Step 2:

[0916] The server periodically obtains the matching status of all users.

[0917] Input: All user matching information stored in the database

[0918] Output: Matching status data for all users obtained by the server

[0919] Specific operation: The server uses a scheduler to retrieve the matching status of all users from the database at regular intervals (e.g., every hour).

[0920] Step 3:

[0921] The matching status data acquired by the server is input into the AI ​​model.

[0922] Input: Matching status data of all users acquired by the server

[0923] Output: Matching situation data input into the AI ​​model

[0924] Specific operation: The server converts the acquired matching status data into an appropriate format and passes it to the AI ​​model in order to input it into the AI ​​model.

[0925] Step 4:

[0926] The AI ​​model identifies users with low matches.

[0927] Input: Matching situation data input into the AI ​​model

[0928] Output: List of users with few matches

[0929] Specific operation: The AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time (e.g., one week).

[0930] Step 5:

[0931] The AI ​​model selects recommended users for the identified user.

[0932] Input: List of users with few matches, user attribute data

[0933] Output: A list of recommended users

[0934] Specific operation: The AI ​​model selects recommended users for the identified user based on their attributes and past matching history.

[0935] Step 6:

[0936] The server notifies the target user of the information about the recommended user.

[0937] Input: List of recommended users

[0938] Output: Information about the recommended user notified to the target user

[0939] Specific operation: The server notifies the target user of the information about the recommended users selected by the AI ​​model. Notifications are sent to the device via push notification or email.

[0940] (Application example 3)

[0941] 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."

[0942] Conventional systems tend to be prone to bias in the amount of matching a user receives, and there was a problem in that appropriate responses were not made to users who had not been matched for a certain period of time. Furthermore, even on online shopping sites, it was difficult to effectively recommend specific products because users' purchase and browsing histories could not be fully utilized. This led to issues such as lower user satisfaction and a decrease in the frequency of site use.

[0943] 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.

[0944] In this invention, the server includes a means for creating a profile picture using AI, a means for inducing potential matches using advanced attribute judgment by AI, a means for reducing bias in the amount of matches, a means for analyzing a user's purchase history and browsing history and recommending specific products, and a means for actively recommending products to users who have not made a purchase for a certain period of time. This makes it possible to reduce bias in the amount of matches for users and effectively recommend specific products.

[0945] A "means for creating a profile picture" is a means for using AI to generate a profile picture based on a user's characteristics and preferences.

[0946] "Advanced attribute judgment" means that AI analyzes a user's behavioral history and profile information to make judgments to recommend the most suitable match.

[0947] "Means to reduce imbalances in the amount of matches" refers to a method in which AI monitors the matching status of all users and actively recommends potential matches to users with few matches.

[0948] "Means for analyzing purchase history and browsing history" refers to a means by which AI analyzes a user's past purchase history and browsing history to understand their interests and concerns.

[0949] "Means of recommending specific products" refers to a means of recommending products that match the user based on the user's purchasing history and browsing history analyzed by AI.

[0950] The "means for actively recommending products to users who have not made a purchase for a certain period of time" refers to a means for identifying users who have not made a purchase for a certain period of time and actively recommending products to those users.

[0951] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The server includes a means for creating profile pictures using AI, a means for determining advanced attributes, a means for reducing bias in the amount of matching, a means for analyzing purchase history and browsing history, a means for recommending specific products, and a means for proactively recommending products to users who have not made a purchase for a certain period of time.

[0952] Server Processing

[0953] The server first collects user profile information and behavioral history, then analyzes this data using AI. Specifically, it uses the pandas library to load user and product data, and uses scikit-learn's cosine_similarity to recommend products based on the user's attributes. This allows it to generate a profile picture based on the user's characteristics and preferences, and recommend optimal matches and products.

[0954] Terminal handling

[0955] The device provides an interface for users to access the service. Through a device such as a smartphone or PC, users can create a profile picture and check the results of product recommendations. In addition, if a user has not made a purchase for a certain period of time, the system will actively recommend products to them via push notifications and emails.

[0956] User operations

[0957] Users enter their profile information through their devices and provide their purchase and browsing history. This allows the server to accurately understand the user's attributes and recommend the most suitable match and products. The user can then check the recommended products and purchase them.

[0958] Specific examples

[0959] For example, if a female user in her 30s is interested in fashion and has previously purchased casual wear and sneakers, the server will analyze the user's purchasing history and recommend the following products:

[0960] Example prompt sentence:

[0961] User ID: 12345

[0962] Attributes: Women in their 30s, fashion enthusiasts, mid-range price range

[0963] Purchase history: casual wear, sneakers, accessories

[0964] Recommended products:

[0965] casual jacket

[0966] Sports shoes

[0967] Simple necklace

[0968] By inputting this prompt sentence into a generative AI model, the most suitable product recommendations are made to the user.

[0969] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0970] Step 1:

[0971] The server reads user data and product data. Specifically, it reads data from a CSV file using the pandas library. The input is the CSV file of user data and product data, and the output is the user data and product data in data frame format.

[0972] Step 2:

[0973] The server analyzes the user's purchase history and browsing history. Specifically, it extracts the purchase history of a specific user and obtains the user's attribute information. The input is the user ID, and the output is the user's purchase history data.

[0974] Step 3:

[0975] The server uses an AI model to recommend products based on user attributes. Specifically, it uses scikit-learn's cosine_similarity to calculate the similarity between the user's attribute vector and the product vector. The input is the user's attribute information and product data, and the output is a list of products with high similarity.

[0976] Step 4:

[0977] The server identifies users who have not made a purchase for a certain period of time. Specifically, it compares the current date with the user's last purchase date and extracts users who have not made a purchase for a certain period of time (e.g., 30 days). The input is user data, and the output is a list of users who have not made a purchase for a certain period of time.

[0978] Step 5:

[0979] The server proactively recommends products to users who have not made a purchase for a certain period of time. Specifically, it uses the product list obtained in step 3 to send push notifications and emails to users. The input is a list of users who have not made a purchase for a certain period of time and a product list, and the output is a notification of recommended products sent to the user.

[0980] Step 6:

[0981] The terminal provides an interface that allows users to check recommended products. Specifically, it displays recommended products on the screen of a smartphone or PC. The input is recommended product data sent from the server, and the output is recommended product information that is displayed to the user.

[0982] Step 7:

[0983] The user checks the recommended products through the terminal and purchases them. Specifically, they click on the recommended products to check detailed information and then complete the purchase procedure. The input is the recommended product information displayed on the terminal, and the output is the completion of the purchase procedure.

[0984] 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.

[0985] "Example 1"

[0986] In one embodiment of the present invention, in addition to a means for creating a profile picture using AI, a means for determining advanced attributes using AI, and a means for reducing bias in the amount of matches, an emotion engine that recognizes the user's emotions is incorporated. This emotion engine recognizes emotions from the user's text input or voice input and recommends matching partners based on those emotions. For example, if a user inputs the text "I'm having fun," the emotion engine recognizes that the user is currently in a happy mood and recommends matching partners that match that mood.

[0987] "Example 2"

[0988] In another embodiment of the present invention, the emotion engine tracks changes in the user's emotions and re-recommends matches based on those changes. For example, if a user first inputs the text "I'm happy" and then later inputs the text "I'm lonely," the emotion engine recognizes that the user's emotions have changed from happy to lonely and re-recommends matches that match the user's new emotions.

[0989] "Example 3"

[0990] Furthermore, in another embodiment of the present invention, the emotion engine learns changes in the user's emotions and recommends matching partners based on the learning results. For example, if a user tends to input the text "fun" every weekend, the emotion engine learns this tendency and recommends matching partners that match the user's happy mood on weekends.

[0991] The processing flow of each embodiment will be described below.

[0992] "Example 1"

[0993] Step 1: The user enters text or voice input into the application.

[0994] Step 2: The emotion engine recognizes emotions from the user's input. For example, if the text "fun" is entered, the emotion engine recognizes that the user is in a happy mood.

[0995] Step 3: Based on the emotions recognized by the emotion engine, the AI ​​recommends potential matches, taking into account the user's profile information and past behavioral history.

[0996] "Example 2"

[0997] Step 1: The user enters text or voice input into the application.

[0998] Step 2: The emotion engine recognizes emotions from the user's input and tracks changes in those emotions. For example, if a user first inputs "happy" and then later inputs "lonely," the emotion engine will recognize the change in emotion.

[0999] Step 3: Based on the changes in emotions recognized by the emotion engine, the AI ​​re-recommends potential matches, taking into account the user's profile information and past behavioral history.

[1000] "Example 3"

[1001] Step 1: The user enters text or voice input into the application.

[1002] Step 2: The emotion engine recognizes emotions from the user's input and learns changes in those emotions. For example, if a user tends to input "fun" every weekend, the emotion engine will learn that tendency.

[1003] Step 3: Based on the emotional trends learned by the emotion engine, the AI ​​recommends potential matches, taking into account the user's profile information and past behavioral history.

[1004] Example 1

[1005] 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."

[1006] In conventional systems, it was difficult to fully reflect the user's characteristics and emotions when generating a user's profile picture or recommending potential matches. Furthermore, since matchmaking recommendations were not based on the user's emotions, there was a risk of user satisfaction decreasing. This led to a need for improved matching accuracy and user experience.

[1007] 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.

[1008] In this invention, the server includes means for inputting user profile information, means for receiving the input information, means for inputting a prompt sentence to the generative AI model, means for the generative AI model to generate a profile picture, means for analyzing the user's emotions using an emotion engine, means for recommending matching partners based on the emotions, and means for the user to confirm the generated profile picture and the recommended matching partners. This makes it possible to generate a profile picture that reflects the user's characteristics and emotions, and to recommend optimal matching partners based on the emotions.

[1009] The "means for inputting user profile information" refers to a device or software that provides an interface for users to input information such as self-introduction, hobbies, and preferences.

[1010] "Means for receiving input information" refers to a device or software that has the functionality to allow the server to receive, store, or process the profile information input by the user.

[1011] "Means for inputting prompt sentences into a generative AI model" refers to a device or software for inputting prompt sentences generated based on a user's profile information into an AI model.

[1012] A "means for a generative AI model to generate a profile picture" is a device or software that uses AI technology to generate a profile picture based on a user's characteristics and preferences.

[1013] The "means for analyzing user emotions using an emotion engine" is a device or software for analyzing emotions from a user's text input or voice input.

[1014] The "means for recommending a match based on emotion" is a device or software for recommending the most suitable match based on the analyzed emotion information of the user.

[1015] The "means for the user to review the generated profile picture and recommended matches" refers to a device or software that provides an interface that allows the user to review the generated profile picture and recommended matches.

[1016] MODE FOR CARRYING OUT THE INVENTION

[1017] This invention is a system that generates a profile picture using a generative AI model based on a user's profile information, analyzes the user's emotions using an emotion engine, and recommends optimal matches. Specific embodiments of this system are described below.

[1018] Program Generation

[1019] The server creates a program to generate a user's profile picture. This program uses a generative AI model to generate a profile picture based on information such as a self-introduction, hobbies, and preferences provided by the user. It also includes a function to analyze the user's emotions using an emotion engine and recommend potential matches based on those emotions.

[1020] Program processing

[1021] The server receives profile information entered by the user through the device. For example, the user might enter information such as "I like the outdoors and often go camping." Based on this information, the server generates a prompt sentence and inputs it into the generative AI model. Image generation models such as "Stable Diffusion" and "DALL-E" can be used as the generative AI model.

[1022] The generative AI model generates a profile picture based on a prompt received from the server. For example, the following prompt is input to the generative AI model:

[1023] The user's hobby is the outdoors, and they like camping. Based on this information, generate a profile picture with an outdoor scene in the background, such as a campsite or a mountain.

[1024] Based on this prompt, the generative AI model generates a profile photo with a campsite as the background.

[1025] Furthermore, the server uses an emotion engine to analyze the user's emotions. If the user inputs "I'm in a happy mood today," the emotion engine analyzes the information and recognizes that the user is in a happy mood. The emotion engine can use text analysis or voice analysis technology.

[1026] The server then recommends a match based on the user's emotional information. For example, if a user is in a happy mood, the server recommends users who are in a similar happy mood or who share fun activities.

[1027] Specific examples

[1028] If the user inputs "I like the outdoors and often go camping," the server will input the following prompt sentence into the generative AI model:

[1029] The user's hobby is the outdoors, and they like camping. Based on this information, generate a profile picture with an outdoor scene in the background, such as a campsite or a mountain.

[1030] Based on this prompt, the generative AI model generates a profile photo with a campsite as the background.

[1031] For example, if a user types "I'm in a happy mood today," the server will use its emotion engine to analyze the user's emotions and recommend a match that matches their happy mood. For example, it will recommend users who also like the outdoors and are in a happy mood.

[1032] In this way, the server can use AI technology based on the information provided by the user to generate profile pictures and recommend potential matches.

[1033] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1034] Program processing flow

[1035] Step 1: User enters profile information

[1036] The user inputs profile information such as a self-introduction, hobbies, and preferences through the terminal. For example, the user may input information such as "I like the outdoors and often go camping."

[1037] Input: User profile information

[1038] Output: Sending profile information from device to server

[1039] Step 2: The server receives the information

[1040] The server receives the profile information sent by the user from the terminal, which is used in subsequent processing.

[1041] Input: Profile information sent from the device

[1042] Output: Profile information stored on the server

[1043] Step 3: The server inputs a prompt to the generative AI model.

[1044] The server generates a prompt based on the received profile information and inputs it into the generative AI model. For example, it generates a prompt such as, "The user's hobby is the outdoors, and they like camping. Based on this information, please generate a profile picture with an outdoor scene, such as a campsite or mountain, as the background."

[1045] Input: Profile information stored on the server

[1046] Output: The prompt sentence that is input to the generative AI model

[1047] Step 4: Generative AI model generates profile photo

[1048] Generative AI models (such as "Stable Diffusion" and "DALL-E") generate profile pictures based on prompts received from the server, with backgrounds that match the user's interests and preferences.

[1049] Input: A prompt sentence entered into the generative AI model

[1050] Output: Generated profile picture

[1051] Step 5: The server uses the emotion engine to analyze the user's emotions.

[1052] The server uses an emotion engine to analyze emotions from text or voice input by the user. For example, if the user inputs "I'm in a happy mood today," the emotion engine analyzes the information and recognizes that the user is in a happy mood.

[1053] Input: User-entered text or speech

[1054] Output: Analyzed user emotion information

[1055] Step 6: The server recommends matches based on emotions

[1056] The server recommends matching partners based on the emotional information analyzed by the emotion engine. For example, if a user is in a happy mood, the server will recommend users who are in a similar happy mood or who can share fun activities.

[1057] Input: Analyzed user emotion information

[1058] Output: A list of suggested matches

[1059] Step 7: User reviews generated profile photo and suggested matches

[1060] Users can view the profile picture generated and the recommended matches through their device, and decide their next action based on this information.

[1061] Input: A generated profile picture and a list of suggested matches

[1062] Output: User confirmation and next action

[1063] (Application example 1)

[1064] 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."

[1065] In conventional systems, the generation of user profile pictures and avatars does not adequately reflect the user's characteristics and preferences, resulting in low user satisfaction. In addition, since matching does not take the user's emotions into consideration, matching accuracy is low, and the user experience is not improved.

[1066] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means. In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches through advanced attribute judgment by AI, means for reducing bias in the amount of matches, means for recognizing a user's emotions and adjusting the facial expression and pose of an avatar based on the emotions, and means for generating an avatar based on the user's characteristics and preferences. This makes it possible to generate a profile picture and avatar based on the user's characteristics and preferences, and furthermore, since matching is performed taking the user's emotions into consideration, it is possible to improve user satisfaction and matching accuracy.

[1067] "Means for creating profile pictures using AI" refers to a means for generating profile pictures using artificial intelligence technology based on information such as a self-introduction, hobbies, and preferences provided by the user.

[1068] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match.

[1069] "Means for inducing potential matches" refers to means for finding and recommending potential matches based on the user's characteristics and preferences.

[1070] "Means for reducing bias in the amount of matching" refers to means for avoiding concentration of matching on specific users and for achieving a balance in overall matching.

[1071] "Means for recognizing a user's emotions and adjusting the avatar's facial expression and pose based on those emotions" refers to means for recognizing emotions from a user's text input or voice input and automatically changing the avatar's facial expression and pose in accordance with those emotions.

[1072] "Means for generating an avatar based on a user's characteristics and preferences" refers to a means for using artificial intelligence technology to generate an avatar that is optimal for a user based on information such as a self-introduction, hobbies, and preferences provided by the user.

[1073] The following describes an embodiment of the present invention: First, the overall system configuration will be described, and then the specific operation of each means will be described.

[1074] System configuration

[1075] This system includes a user terminal, a server, and an AI model. The user terminal provides an interface for the user to input information such as a self-introduction, hobbies, and preferences. The server receives this information and generates a profile picture and avatar using the AI ​​model. The server also uses an emotion engine to recognize the user's emotions and adjust the avatar's facial expression and pose based on those emotions.

[1076] Program processing

[1077] The server receives the information provided by the user and performs the following processes:

[1078] 1. Generate profile photo:

[1079] The server uses a generative AI model to generate a profile picture based on information entered by the user, such as a self-introduction, hobbies, preferences, etc. For example, if a user provides information that they "like the outdoors," the AI ​​model will generate a profile picture with an outdoor scene, such as a campsite or mountain, as the background.

[1080] 2. Avatar generation:

[1081] The server generates an avatar based on the user's characteristics and preferences, and the generated avatar is optimized for the information provided by the user.

[1082] 3. Emotion Recognition and Facial Expression Modulation:

[1083] The server uses an emotion engine that recognizes emotions from the user's text and voice input. For example, if the user inputs the text "I'm having fun," the emotion engine recognizes that the user is currently in a happy mood and adjusts the avatar's facial expression and pose based on that emotion.

[1084] Hardware and software used

[1085] Hardware:

[1086] User devices: smartphones, tablets, computers, etc.

[1087] Server: A server with high-performance computing power

[1088] software:

[1089] Generative AI models: OpenAI APIs, etc.

[1090] Emotion Engine: Emotion recognition engine using natural language processing technology

[1091] Image processing library: PIL (Python Imaging Library) etc.

[1092] Specific examples

[1093] If the user provides information that "I like the outdoors" and also inputs the emotion "fun," the following prompt sentence is generated:

[1094] Prompt Sentence Examples

[1095] Generate an avatar with an outdoor scene background based on the user's bio: I love the outdoors, and my hobbies are camping and hiking.

[1096] By inputting this prompt into a generative AI model, an avatar based on the user's characteristics and preferences is generated. In addition, if the emotion engine recognizes the user's emotion as "happy," the avatar's facial expression is adjusted to a smile.

[1097] In this way, it becomes possible to generate profile pictures and avatars based on the user's characteristics and preferences, and matching is performed taking the user's emotions into consideration, thereby improving user satisfaction and matching accuracy.

[1098] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1099] Step 1:

[1100] The user uses the device to input information such as a self-introduction, hobbies, and preferences. The input information is then sent from the device to the server. Examples of input data include "I like the outdoors," "camping," and "hiking."

[1101] Step 2:

[1102] The server generates a prompt based on the received user information. The generated prompt is then input into the generative AI model. Specifically, the following prompt is generated:

[1103] Generate an avatar with an outdoor scene background based on the user's bio: I love the outdoors, and my hobbies are camping and hiking.

[1104] By inputting this prompt into a generative AI model, an avatar based on the user's characteristics and preferences is generated.

[1105] Step 3:

[1106] The generative AI model generates an avatar based on the prompt text. The image data of the generated avatar is returned to the server. The server receives this image data and proceeds to the next processing step.

[1107] Step 4:

[1108] The server uses an emotion engine to recognize emotions from the user's text input or voice input. For example, if the user inputs the text "fun," the emotion engine recognizes that the user is currently in a happy mood. The input data is the text "fun," and the output data is the emotion information "fun."

[1109] Step 5:

[1110] The server adjusts the facial expression and pose of the generated avatar based on the emotional information obtained from the emotion engine. For example, if the user's emotion is recognized as "fun," the server adjusts the avatar's facial expression to a smile. The input data is the emotional information "fun," and the output data is an avatar image with a smiling face.

[1111] Step 6:

[1112] The server sends the final adjusted avatar image to the user's device, where the user can view the generated avatar and make further adjustments as needed. The input data is the adjusted avatar image, and the output data is the avatar image displayed on the user's device.

[1113] The above processing steps make it possible to generate profile pictures and avatars based on the user's characteristics and preferences, and also to perform matching that takes the user's emotions into consideration, thereby improving user satisfaction and matching accuracy.

[1114] Example 2

[1115] 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."

[1116] Conventional systems were unable to fully utilize the user's behavioral history and profile information, making it difficult to recommend optimal matches. Furthermore, because the system did not re-recommend matches based on the user's emotional changes, it was unable to increase user satisfaction. This made it difficult for users to continue using the system, and the success rate of matches also decreased.

[1117] 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.

[1118] In this invention, the server includes means for collecting user behavior history and profile information, means for preprocessing the collected data, means for inputting the preprocessed data into an AI model and recommending optimal matching partners, means for analyzing text entered by the user and tracking changes in emotions, and means for re-recommending matching partners based on the changes in emotions. This makes it possible to recommend optimal matching partners based on the user's behavior history and profile information, and furthermore, to perform re-recommendations taking into account changes in the user's emotions, thereby increasing user satisfaction and improving the success rate of matching.

[1119] "User behavior history" is a record of a series of operations and activities performed by a user within the system.

[1120] "Profile information" refers to information such as personal characteristics, preferences, and hobbies provided by a user.

[1121] "Preprocessing" is a series of operations that convert collected data into a format that is easy to analyze.

[1122] An "AI model" is an algorithm or system that uses artificial intelligence to analyze data and perform specific tasks.

[1123] A "matching partner" is another user who is recommended as a suitable partner for the user.

[1124] "Text analysis" is the process of analyzing the text entered by the user and understanding its content and sentiment.

[1125] "Changes in emotion" refers to fluctuations in emotion that can be read from the text entered by the user.

[1126] "Re-recommendation" refers to recommending a match partner again based on new information or conditions after the initial recommendation.

[1127] MODE FOR CARRYING OUT THE INVENTION

[1128] This invention is a system that collects user behavior history and profile information, and uses an AI model to recommend optimal matches based on that information. It also includes a function to track changes in the user's emotions and re-recommend matches based on those changes.

[1129] 1. Generate the system program

[1130] The server collects the user's behavioral history and profile information and stores it in a database. This includes information about the user's hobbies and preferences, and the people with whom they frequently communicate. The server inputs this data into an AI model and recommends the user the best possible match.

[1131] 2. Explain the generated program's processing in natural language

[1132] The server uses Python and machine learning libraries such as TensorFlow and PyTorch to analyze user data. For example, if a user's hobby is "watching movies" and they frequently exchange messages with other users who share the same hobby, the AI ​​will learn this tendency and recommend users who also share the hobby as potential matches.

[1133] The emotion engine also analyzes the text entered by the user and tracks changes in emotion. The server uses a natural language understanding API as the emotion engine. For example, if a user enters "I'm happy today" and then later enters "I'm lonely today," the emotion engine will recognize this change and re-recommend matches that match the user's new emotions.

[1134] 3. Add examples to your description

[1135] As a concrete example, assume that User A has hobbies of "watching movies" and "reading" and frequently communicates with other users who share these hobbies. The server collects User A's behavioral history and profile information and inputs it into an AI model. The AI ​​then recommends User B, whose hobbies are "watching movies" and "reading," as a potential match for User A.

[1136] For example, if User A inputs "I'm having fun today," and then a short time later inputs "I feel lonely today," the emotion engine will recognize this change. The server will then re-recommend User C, who also feels lonely, as a match that matches User A's new emotion.

[1137] Examples of prompts to input to a generative AI model might include:

[1138] Please recommend the best match based on the behavioral history and profile information of User A. User A's hobbies are "watching movies" and "reading," and he frequently communicates with other users who share these hobbies.

[1139] In this way, the processing of the system is explained while clarifying how the server, terminal, and user are involved.

[1140] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1141] Step 1:

[1142] User Data Collection

[1143] The server collects profile information and behavioral history provided by the user. Specifically, it stores the hobbies and preferences entered by the user, message exchanges, login history, etc. in a database. Input is performed by the user via a terminal, and output is user data stored on the server.

[1144] Step 2:

[1145] Data Preprocessing

[1146] The server preprocesses the collected data. Specifically, it performs tasks such as imputing missing values, normalizing data, and encoding categorical data. For example, it converts a user's hobbies, "watching movies" and "reading books," into numerical data. The input is the collected raw data, and the output is the preprocessed data.

[1147] Step 3:

[1148] Analysis and matchmaking recommendations using AI models

[1149] The server inputs the preprocessed data into the AI ​​model. The AI ​​model is built using TensorFlow and PyTorch. The model recommends optimal matches based on the user's behavioral patterns and profile information. For example, if user A's hobby is "watching movies," the model recommends user B who has the same hobby. The input is the preprocessed data, and the output is a list of recommended matches.

[1150] Step 4:

[1151] Emotion Tracking with Emotion Engine

[1152] The server sends the text entered by the user to the emotion engine. The emotion engine uses a natural language understanding API to analyze the user's emotion. For example, if the user enters "I'm having fun today," the emotion engine recognizes the emotion as "fun." The input is the text entered by the user, and the output is the analyzed emotion data.

[1153] Step 5:

[1154] Emotion-based re-recommendation

[1155] The server re-recommends matching partners based on the emotional data obtained from the emotion engine. For example, if a user inputs "I'm having fun today" and then "I'm lonely today," the server recognizes the change in emotion and re-recommends matching partners that fit the new emotion. The input is the emotional data from the emotion engine, and the output is a list of re-recommended matching partners.

[1156] (Application example 2)

[1157] 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."

[1158] Conventional systems match users based on their behavioral history and profile information, but do not consider changes in the user's emotions when matching or recommending content. As a result, they are unable to provide optimal matches or content that match the user's emotions, making it difficult to improve user satisfaction. Furthermore, there are also insufficient means to reduce bias in the amount of matching.

[1159] 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.

[1160] In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches using advanced attribute judgment by AI, means for reducing bias in the amount of matches, means for tracking changes in the user's emotions and re-recommending matches based on those changes, and means for analyzing the user's behavior history and profile information and recommending optimal content, thereby making it possible to provide optimal matches and content that match the user's emotions.

[1161] "Means for creating profile pictures using AI" refers to a means for generating profile pictures using artificial intelligence based on the user's characteristics and preferences.

[1162] "Advanced attribute judgment using AI" is a method of analyzing a user's behavioral history and profile information and using artificial intelligence to make judgments to recommend the most suitable match.

[1163] "Means for inducing potential matches" refers to means for finding potential matches based on the user's attributes and behavioral history and recommending them to the user.

[1164] "Means for reducing bias in the amount of matching" refers to means for reducing bias in matching for specific users and achieving a balance in overall matching.

[1165] The "means for tracking changes in a user's emotions" is a means for detecting changes in emotions from the text entered by the user or from the user's actions, and recording those changes.

[1166] The "means for re-recommending matching partners based on emotions" refers to a means for re-recommending appropriate matching partners based on changes in the user's emotions.

[1167] "Means for analyzing a user's behavior history and profile information" refers to means for collecting a user's past behavior and profile information and conducting analysis based on that information.

[1168] The "means for recommending optimal content" is a means for recommending optimal content to a user based on the user's behavioral history and changes in emotions.

[1169] A system for implementing this invention includes a server, a user terminal, and an AI model. The server collects user behavior history and profile information and performs advanced attribute judgments based on this data. Furthermore, the server tracks changes in the user's emotions and re-recommends matching partners and content based on those changes.

[1170] 1. System Program

[1171] The server runs a program that includes the following main means:

[1172] How profile pictures are created: Profile pictures are generated using AI based on user characteristics and preferences.

[1173] Advanced attribute determination methods: Analyzes user behavior history and profile information to recommend optimal matches.

[1174] Measures to reduce bias in matching volume: Reduce bias in matching towards specific users.

[1175] A means of tracking emotional changes: Detecting emotional changes from the text entered by the user and their actions, and recording those changes.

[1176] A method for re-recommending matching partners based on emotions: Based on changes in the user's emotions, appropriate matching partners are re-recommended.

[1177] A means of recommending optimal content: Recommending optimal content to users based on their behavioral history and emotional changes.

[1178] 2. Program processing explanation

[1179] The server implements these methods using a programming language such as Python. Specifically, it performs the following processes:

[1180] Profile picture creation: Using an AI model (e.g., GAN) to generate a profile picture based on the user's characteristics and preferences as input.

[1181] Advanced attribute determination: User behavioral history and profile information are retrieved from a database and analyzed using an AI model (e.g., TensorFlow or PyTorch).

[1182] Reduce bias in match volume: Adjust the matching algorithm to reduce bias in matches against certain users.

[1183] Sentiment change tracking: Analyzes user-entered text using natural language processing (NLP) techniques (e.g., BERT) to detect emotional changes.

[1184] Emotion-based re-recommendation: Based on changes in emotions, the AI ​​model is used again to recommend suitable matches.

[1185] Optimal content recommendation: Recommend optimal content using an AI model based on the user's behavioral history and emotional changes.

[1186] 3. Examples and prompts

[1187] As a concrete example, consider a scenario in which an application recommends a "Comedy Show" after a user inputs the emotion "fun." In this case, the server inputs the following prompt sentence into the generative AI model:

[1188] Example prompt sentence:

[1189] The user entered the emotion "fun." Please recommend content that matches their current emotion.

[1190] Using this prompt, the AI ​​model can recommend the most appropriate content based on the user's emotions.

[1191] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1192] Step 1:

[1193] The server collects user behavior history and profile information. It retrieves data entered by the user when using the application and past behavior history from the database. This provides input data for understanding the user's characteristics and preferences.

[1194] Step 2:

[1195] The server uses AI models to make sophisticated attribute judgments based on the collected user behavior history and profile information. Specifically, it uses machine learning frameworks such as TensorFlow and PyTorch to analyze the user's characteristics and recommend the most suitable matchmaking partners. This process outputs a list of the most suitable matches for the user.

[1196] Step 3:

[1197] The server tracks changes in the user's emotions. It analyzes the text data entered by the user using natural language processing (NLP) techniques (e.g., BERT) to detect changes in emotions. The input data is the text entered by the user in the application, and the output is the detected changes in emotions.

[1198] Step 4:

[1199] The server then uses the AI ​​model again to recommend suitable matches based on the changes in emotions. Using the changes in emotions as input data, the matching algorithm is run again to output matches that match the user's current emotions.

[1200] Step 5:

[1201] The server recommends optimal content based on the user's behavioral history and emotional changes. Using the user's behavioral history and emotional changes as input data, the server uses an AI model to select optimal content from a content database and recommend it to the user. This process outputs a list of content that is optimal for the user.

[1202] Step 6:

[1203] The user terminal displays a list of optimal matches and content received from the server. The user can select from the displayed matches and content and take the next action. This improves user satisfaction.

[1204] Example 3

[1205] 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."

[1206] In conventional systems, there was a tendency for bias in the amount of matching between users, which made it difficult for certain users to find a match. Furthermore, since the system did not take into account the user's emotions or mood when recommending a match, there was a possibility that user satisfaction would decrease. There is a need to solve these issues and provide a fairer and more satisfying matching experience.

[1207] 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.

[1208] In this invention, the server includes means for generating profile images using AI, means for inducing potential matches through advanced attribute analysis using AI, means for reducing bias in the amount of matches, means for collecting user emotion data and analyzing it with an emotion engine, and means for recommending matches based on the analysis results of the emotion engine. This reduces bias in the amount of matches for users and enables recommendation of matches based on emotions.

[1209] "AI-based profile image generation means" is a function that uses artificial intelligence to automatically generate profile images based on the user's characteristics and preferences.

[1210] "Advanced attribute analysis using AI" is the process of using artificial intelligence to analyze a user's behavioral history and profile information to identify the most suitable match.

[1211] "Means to induce potential matches" is a function in which artificial intelligence recommends potential compatible match partners based on the user's attributes and behavioral data.

[1212] "Means to reduce imbalances in the amount of matches" refers to a function that monitors the matching status of all users and actively recommends potential matches to users with few matches.

[1213] The "means for collecting user emotional data" is a function for collecting text data and behavioral data entered by the user and understanding the user's emotional state.

[1214] "Means for analyzing with an emotion engine" refers to the process of analyzing collected user emotion data using an emotion engine and learning changes in the user's emotions.

[1215] The "means for recommending a match partner based on the analysis results of the emotion engine" is a function that recommends a match partner that matches the user's emotional state based on the analysis results of the emotion engine.

[1216] MODE FOR CARRYING OUT THE INVENTION

[1217] This invention is a system that uses AI to generate profile images, performs advanced attribute analysis, reduces bias in the amount of matches, and recommends matching partners based on emotions using an emotion engine, in order to improve users' matching experience.

[1218] Hardware and software used

[1219] The server uses the following hardware and software:

[1220] Hardware: High-performance server machine (CPU, GPU, memory, storage)

[1221] Software: TensorFlow, PyTorch, database management system (DBMS), mail server, push notification service, Google's Natural Language API, IBM Watson sentiment analysis tool

[1222] Terminals are devices used by users (smartphones, tablets, PCs) that use the following software:

[1223] Software: Chat app, diary app, data collection module

[1224] Program processing

[1225] The server periodically collects the matching history data of all users and identifies users who have not had a match for a certain period of time (for example, one month). For these identified users, AI is used to recommend potential matches. The AI ​​model used here is built using TensorFlow and PyTorch.

[1226] Examples:

[1227] If user A has no matches for a month, the server will recommend users B and C based on user A's attributes (age, hobbies, region, etc.).

[1228] The device collects text data entered by the user, for example, by saving text entered through a chat app or diary app, and periodically transmits the text to a server.

[1229] The server analyzes the user's text data and learns changes in emotions using an emotion engine, which uses Google's Natural Language API and IBM Watson's emotion analysis tools.

[1230] Examples:

[1231] If user D tends to input the text "fun" every weekend, the server learns this tendency and recommends user E, who matches the "fun" mood, to user D on the weekend.

[1232] Prompt Sentence Examples

[1233] Here are some example prompts to input to the generative AI model:

[1234] "For users who have not had a match for a month, please recommend users who are likely to match their attributes."

[1235] "If a user tends to type 'fun' on the weekends, recommend matches that fit that user."

[1236] In this way, the system can improve the matching experience of the user. The flow of the identification process in the third embodiment will be described with reference to FIG.

[1237] Program processing flow

[1238] Step 1: Monitor your users' matches

[1239] The server periodically collects the matching history data of all users. As input, it retrieves the matching history of each user from the database. As output, it generates a list of users who have not had a match for a certain period of time (e.g., one month).

[1240] Specific behavior:

[1241] The server runs a script every night at midnight to check the matching history of all users.

[1242] The server lists users who have not had a match for more than a month.

[1243] Step 2: Make recommendations for users with few matches

[1244] The server uses AI to recommend potential matches for the identified user. The input is the user's attribute data (age, hobbies, region, etc.) and is input into the AI ​​model. The output is a list of recommended users.

[1245] Specific behavior:

[1246] The server inputs the identified user's attribute data into the AI ​​model.

[1247] The server obtains the recommended user list output from the AI ​​model and stores it in a database.

[1248] Step 3: Notify the user of the recommendation

[1249] The server notifies the user of the recommendation results. As input, it generates a notification message based on the recommended user list. As output, it sends a notification to the user via email or in-app notification.

[1250] Specific behavior:

[1251] The server generates a notification message based on the recommended user list.

[1252] The server sends notifications to the user through a mail server or push notification service.

[1253] Step 4: Collect user text data

[1254] The device collects text data entered by the user. As input, it stores the text entered through a chat app or diary app. As output, it sends the collected text data to a server.

[1255] Specific behavior:

[1256] The device captures text entered by the user in real time and stores it in local storage.

[1257] The terminal periodically transmits text data to the server.

[1258] Step 5: Learn emotional changes with the emotion engine

[1259] The server analyzes the user's text data using an emotion engine and learns changes in emotions. The received text data is input to the emotion engine. As an output, the emotion analysis results are generated and stored in a database.

[1260] Specific behavior:

[1261] The server inputs the received text data into the emotion engine.

[1262] The server stores the emotion analysis results output from the emotion engine in a database.

[1263] Step 6: Recommend matches based on emotions

[1264] The server recommends a match that matches the user's emotions based on the learning results of the emotion engine. As input, it identifies the user's emotional patterns based on the emotion analysis results. As output, it uses an AI model to recommend other users with similar emotional patterns and generates a recommendation list.

[1265] Specific behavior:

[1266] The server identifies the user's emotional pattern based on the emotion analysis results.

[1267] The server uses an AI model to recommend other users with similar emotional patterns and generate a recommendation list.

[1268] The server notifies the user of the recommendation list.

[1269] (Application example 3)

[1270] 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."

[1271] In conventional systems, there is often a bias in the amount of matching between users, making it difficult for certain users to find a match. In addition, because recommendations do not utilize individual data such as user emotions or order history, it is difficult to provide the optimal match or recommendation for the user. This leads to issues such as lower user satisfaction and a low rate of continued use of the system.

[1272] 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.

[1273] In this invention, the server includes means for creating profile pictures using AI, means for inducing potential matches using advanced attribute judgments by AI, means for reducing bias in the amount of matches, means for making optimal recommendations based on the user's order history and emotional data, and means for the emotion engine to learn changes in the user's emotions and make recommendations based on the learning results. This enables optimal matching and recommendations using individual user data, improving user satisfaction and increasing the rate at which the system is used.

[1274] "AI-based profile picture creation method" is a function that uses artificial intelligence to generate profile pictures based on the user's characteristics and preferences.

[1275] "Advanced attribute judgment using AI" is a function that analyzes a user's behavioral history and profile information and uses artificial intelligence to recommend the most suitable match.

[1276] "Means to reduce imbalances in the amount of matches" refers to a function that monitors the matching status of all users and actively recommends potential matches to users with few matches.

[1277] "Means for making optimal recommendations based on a user's order history and emotional data" is a function that analyzes a user's past order history and emotional data and recommends optimal restaurants and menus based on that.

[1278] "A means for the emotion engine to learn changes in the user's emotions and make recommendations based on the learning results" is a function that learns the user's emotional data and makes optimal matches and recommendations based on those changes.

[1279] To implement the present invention, the following system configuration and program are required.

[1280] System Configuration

[1281] 1. Server: The server generates a user profile picture and runs an AI model to analyze the user's behavioral history and emotional data. The server accesses the database to obtain the user's order history and emotional data, and then makes optimal recommendations based on that.

[1282] 2. User Device: User devices include smartphones, smart glasses, head-mounted displays, or robots. These devices receive the recommendation results from the server and display them to the user.

[1283] Program processing explanation

[1284] The server performs the following process.

[1285] 1. Data collection and storage: The server stores user profile information, behavioral history, order history, and emotion data in a database, including text data entered by users and order history data.

[1286] 2. Profile photo generation: The server generates a profile photo based on the user's characteristics and preferences using an AI model. This AI model generates the optimal photo based on the user's input data.

[1287] 3. Advanced attribute judgment: The server analyzes the user's behavioral history and profile information and recommends the best match using an AI model that identifies potential matches based on the user's past behavioral data.

[1288] 4. Mitigating the imbalance in the amount of matches: The server monitors the matching status of all users and proactively recommends potential matches to users with few matches, thereby reducing the imbalance in the amount of matches.

[1289] 5. Emotional Data Learning and Recommendation: The server uses an emotion engine to learn changes in the user's emotions and makes optimal recommendations based on the learning results. For example, if a user tends to feel "happy" on weekends, the server will recommend restaurants and menus that match the user's happy mood on weekends.

[1290] Specific examples

[1291] As a specific example, consider the case of recommending a restaurant to a user with a user ID of 1. Assume that this user tends to have the emotion "fun" on weekends.

[1292] Example prompt sentence:

[1293] For user ID 1, if the user tends to have the emotion "fun" on weekends, recommend the best restaurant.

[1294] By inputting this prompt into a generative AI model, restaurant recommendations are made based on the user's emotional data.

[1295] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1296] Step 1:

[1297] Data collection and storage

[1298] The server collects user profile information, behavioral history, order history, and emotional data and stores them in a database. Specifically, it acquires text data entered by the user into the application and order history data, and stores this data in the database. The input is the user's profile information, order history, and emotional data, and the output is this data stored in the database.

[1299] Step 2:

[1300] Generate a profile photo

[1301] The server generates a profile picture using an AI model based on the user's characteristics and preferences. Specifically, the server inputs the user's input data (e.g., favorite colors and styles) into the AI ​​model to generate an optimal profile picture. The input is the user's characteristics and preferences, and the output is the generated profile picture.

[1302] Step 3:

[1303] Advanced attribute judgment

[1304] The server analyzes the user's behavioral history and profile information and uses an AI model to recommend optimal matches. Specifically, the user's past behavioral data is input into the AI ​​model to identify potential matches. The input is the user's behavioral history and profile information, and the output is recommended matches.

[1305] Step 4:

[1306] Reducing bias in matching volume

[1307] The server monitors the matching status of all users and proactively recommends new matches to users with few matches. Specifically, it analyzes the matching data of all users, identifies users with few matches, and recommends new matches to those users. The input is the matching data of all users, and the output is the recommended matches.

[1308] Step 5:

[1309] Emotional Data Learning and Recommendation

[1310] The server uses an emotion engine to learn changes in the user's emotions and makes optimal recommendations based on the learning results. Specifically, the user's emotional data is input into the emotion engine, which learns changes in emotions. Then, based on the learning results, the server recommends the optimal restaurant and menu for the user. The input is the user's emotional data, and the output is the recommended restaurant and menu.

[1311] Step 6:

[1312] Displaying recommendation results

[1313] The user terminal receives the recommendation results from the server and displays them to the user. Specifically, the recommendation results sent from the server are received by the user terminal and displayed to the user through the application. The input is the recommendation results from the server, and the output is the recommendation results displayed to the user.

[1314] 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.

[1315] 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.

[1316] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1317] 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.

[1318] [Third embodiment]

[1319] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1320] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1321] 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).

[1322] 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.

[1323] 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.

[1324] The camera 42 includes an optical system including a lens, an aperture, and a shutter, and a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[1325] It is a small digital camera equipped with an imaging element such as a range sensor, and captures an image of the surroundings of the user 20 (for example, an imaging range defined by an angle of view equivalent to the width of the field of vision of a typical healthy person).

[1326] 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.

[1327] 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.

[1328] 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.

[1329] 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.

[1330] 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.

[1331] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1332] "Example 1"

[1333] The system of the present invention utilizes AI technology to create a profile picture for a user. Specifically, the AI ​​generates the profile picture based on information such as a self-introduction, hobbies, and preferences provided by the user. For example, if a user provides information that they "like the outdoors," the AI ​​will set the background of the user's profile picture to an outdoor scene, such as a campsite or mountain.

[1334] "Example 2"

[1335] The system of the present invention also uses AI to perform advanced attribute judgments and guide potential matches. Specifically, the AI ​​analyzes the user's behavioral history and profile information to recommend optimal matches. For example, if a user tends to frequently communicate with people who share specific hobbies or preferences, the AI ​​can learn that tendency and recommend people with similar hobbies and preferences as potential matches.

[1336] "Example 3"

[1337] Furthermore, the system of the present invention reduces the bias in the amount of matches. Specifically, the AI ​​monitors the matching status of all users and more proactively recommends potential matches to users with few matches. For example, if a user has not had a match for a certain period of time, the AI ​​will proactively recommend users who are likely to match the user's attributes.

[1338] The processing flow of each embodiment will be described below.

[1339] "Example 1"

[1340] Step 1: The user provides information such as a self-introduction, hobbies, and preferences.

[1341] Step 2: The AI ​​generates a profile picture based on the provided information. For example, if a user provides information that they "like the outdoors," the AI ​​will set the user's profile picture with an outdoor scene, such as a campsite or mountain, as the background.

[1342] Step 3: The generated profile picture is set as the user's profile.

[1343] Step 1: The user's behavioral history and profile information are analyzed by AI.

[1344] Step 2: The AI ​​recommends optimal matches based on the analysis results. For example, if a user tends to frequently communicate with people who share certain hobbies or preferences, the AI ​​will learn that tendency and recommend people with similar hobbies and preferences as potential matches.

[1345] "Example 3"

[1346] Step 1: AI monitors the matching status of all users.

[1347] Step 2: For users with few matches, the AI ​​will more proactively recommend potential matches. For example, if a user has not had a match for a certain period of time, the AI ​​will proactively recommend users who are likely to match the user's attributes.

[1348] Example 1

[1349] 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."

[1350] In conventional systems, users' profile pictures were based on self-reporting, and often did not fully reflect the user's characteristics and preferences. Furthermore, it was difficult to consider the user's latent attributes and behavioral history when matching, resulting in low matching accuracy. Furthermore, there was also the issue of matching being biased toward certain users, resulting in an imbalance in the overall amount of matches.

[1351] 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.

[1352] In this invention, the server includes means for receiving profile information entered by a user, means for generating a prompt sentence based on the received profile information, means for inputting the generated prompt sentence into a generative AI model and generating a profile picture, means for sending the generated profile picture to the user's device, means for guiding potential matches through advanced attribute judgment by AI, and means for reducing bias in the amount of matches. This makes it possible to generate profile pictures that reflect the user's characteristics and preferences, improving matching accuracy and balancing the overall amount of matches.

[1353] The "means for receiving profile information entered by the user" is a function that allows the server to receive information such as a self-introduction, hobbies, and preferences entered by the user through the terminal.

[1354] "Means for generating a prompt sentence based on received profile information" is a function that creates a prompt sentence to be input into the generative AI model based on the user's profile information received by the server.

[1355] "Means for inputting the generated prompt sentence into a generative AI model and generating a profile picture" refers to a function that inputs the prompt sentence generated by the server into a generative AI model, thereby generating a profile picture that matches the user's characteristics and preferences.

[1356] "Means for sending the generated profile picture to the user's terminal" is a function that allows the server to send the profile picture obtained from the generative AI model to the user's terminal.

[1357] "Means of guiding potential matches through advanced attribute judgment by AI" is a function in which AI analyzes a user's behavioral history and profile information, determines the user's potential attributes, and recommends the most suitable match.

[1358] "Means for reducing bias in the amount of matching" is a function for balancing the overall amount of matching so that matching is not biased towards specific users.

[1359] MODE FOR CARRYING OUT THE INVENTION

[1360] The present invention provides a system for generating a profile picture based on a user's characteristics and preferences, thereby improving the accuracy of matching. Specific embodiments of this system will be described below.

[1361] Program Generation

[1362] The server generates a program to generate a profile picture for the user. This program uses a generative AI model to create the profile picture based on the information provided by the user. Examples of generative AI models that can be used include OpenAI's DALL-E and MidJourney.

[1363] Program processing

[1364] The server receives profile information entered by the user through the terminal. This profile information includes a self-introduction, hobbies, preferences, etc. For example, if a user enters "I like the outdoors," this information is sent to the server.

[1365] Next, the server generates a prompt based on the received profile information. The prompt is text to be input into the generative AI model, and reflects the user's characteristics and preferences. For example, a prompt might be generated that reads, "Please generate a profile picture for a user who loves the outdoors. Please set the background to an outdoor scene, such as a campsite or mountain."

[1366] The server inputs the generated prompt sentence into the generative AI model, which then generates a profile picture that matches the user's characteristics and preferences based on the prompt sentence. The generated profile picture is then sent to the server.

[1367] The server sends the generated profile picture to the user's device, and the user can check the received profile picture on the device and use it as their profile.

[1368] Specific examples

[1369] For example, if a user enters the information "I like reading," the following prompt will be generated:

[1370] "Generate a profile picture for a user who likes reading. Set the background to a reading scene, such as a library or bookshelf."

[1371] Based on this prompt, the generative AI model generates a profile picture with a library and bookshelves as the background, and the generated picture is sent from the server to the user's device, where the user can use the picture as their profile picture.

[1372] In this way, by generating a profile picture based on the user's characteristics and preferences, the accuracy of matching can be improved. Furthermore, advanced attribute judgment using AI can analyze the user's behavioral history and profile information to recommend the most suitable match. Furthermore, by including a means to reduce bias in the amount of matches, the overall amount of matches can be balanced.

[1373] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1374] Step 1:

[1375] The user enters profile information.

[1376] The user uses the terminal to input profile information such as a self-introduction, hobbies, preferences, etc. For example, the user may input "I like the outdoors." This input information is sent to the server.

[1377] Step 2:

[1378] The server receives the user's input information.

[1379] The server receives the profile information sent by the user from the terminal. This information is text data that reflects the user's hobbies and preferences. The received data is used in the next processing step.

[1380] Step 3:

[1381] The server generates a prompt sentence based on the received profile information.

[1382] The server creates a prompt to input to the generative AI model based on the received user profile information. For example, it generates a prompt such as, "Please generate a profile picture for a user who likes the outdoors. Please set the background to an outdoor scene, such as a campsite or mountain." This prompt is used in the next step.

[1383] Step 4:

[1384] The prompt sentence generated by the server is input into the generative AI model.

[1385] The server inputs the generated prompt into a generative AI model (such as OpenAI's DALL-E or MidJourney), which then generates a profile picture that matches the user's characteristics and preferences based on the prompt. The generative AI model then analyzes the prompt and generates an image with an appropriate background and scene.

[1386] Step 5:

[1387] The server receives the generated profile photo.

[1388] The server receives a profile picture generated from the generative AI model, which includes a background and scene that matches the user's interests and preferences. The received profile picture is used in the next step.

[1389] Step 6:

[1390] The server sends the profile picture to the user's device.

[1391] The server sends the received profile picture to the user's device, where the user can use the picture as their profile. The sent picture is displayed on the user's device.

[1392] Step 7:

[1393] The user checks their profile picture.

[1394] The user can review the received profile photo on their device, check if the photo meets their expectations, and request a regeneration if necessary.

[1395] (Application example 1)

[1396] 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."

[1397] In conventional systems, users' profile pictures are based on self-reporting and do not reflect their hobbies or preferences, which means that they cannot fully express the user's personality and interests. Another issue is that poor quality profile pictures result in a lower success rate for matching. Furthermore, because the system does not fully utilize the user's behavioral history and profile information, it is difficult to recommend optimal matches.

[1398] 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.

[1399] In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches through advanced attribute judgment by AI, means for reducing bias in the amount of matches, means for generating prompt sentences based on the user's hobbies and preferences and generating a profile picture using a generative AI model, and means for reflecting the generated profile picture in the user's profile. This makes it possible to generate a high-quality profile picture that reflects the user's personality and interests and recommend optimal matches.

[1400] An "AI profile picture generation method" is a method that uses artificial intelligence technology to generate a profile picture based on user-provided characteristics and preferences.

[1401] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match.

[1402] "Means for inducing potential matches" are means for finding potentially compatible partners based on the user's attributes and behavioral history, and promoting matching.

[1403] "Means for reducing bias in the amount of matching" refers to means for evenly distributing the overall amount of matching so that matching is not concentrated on specific users.

[1404] "Means for generating prompt sentences based on the user's hobbies and preferences" refers to means for creating prompt sentences to be input into a generative AI model based on information about hobbies and preferences provided by the user.

[1405] "Means for generating a profile picture using a generative AI model" refers to means for generating a profile picture for a user using a generative AI model based on a generated prompt sentence.

[1406] The "means for reflecting the generated profile picture in the user's profile" refers to a means for adding the generated profile picture to the user's profile information and displaying it.

[1407] The system for implementing this invention is composed of three main elements: a server, a terminal, and a user. The specific operations of each element and their cooperation will be explained below.

[1408] server

[1409] The server includes a means for creating a profile picture using AI, a means for determining advanced attributes using AI, a means for guiding potential matches, a means for reducing bias in the amount of matches, a means for generating prompt sentences based on the user's hobbies and preferences, a means for generating a profile picture using a generative AI model, and a means for reflecting the generated profile picture in the user's profile.

[1410] The server receives the user's hobbies and preferences and generates a prompt based on them. Using the generated prompt, it uses a generative AI model (e.g., OpenAI's API) to generate a profile picture for the user. The generated profile picture is added to the user's profile information and displayed.

[1411] Terminal

[1412] The terminal is a device where the user enters information and checks the generated profile picture. It can be a smartphone or a PC. The user enters information about their hobbies and preferences through the terminal and sends it to the server. The generated profile picture can be checked on the terminal.

[1413] User

[1414] A user is an individual who uses the system. The user inputs information about their hobbies and preferences through their device and sends it to the server. The server generates a profile picture, which the user can then view on their device.

[1415] Specific examples

[1416] For example, if the user provides information such as "I like fashion," the server generates the following prompt sentence:

[1417] Prompt Sentence Examples

[1418] The user's hobby is fashion, and their preference is stylish cityscapes.

[1419] Based on this prompt, a generative AI model is used to generate a profile picture featuring a fashion show or stylish cityscape as a background, which is then added to the user's profile information and can be viewed on the device.

[1420] In this way, it is possible to generate high-quality profile pictures that reflect the user's personality and interests, and to recommend optimal matches.

[1421] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1422] Step 1:

[1423] The user uses the terminal to input information about his or her hobbies and preferences.

[1424] Input: User's hobbies and preferences (e.g., "I like fashion")

[1425] Output: The entered information is sent to the server.

[1426] Specific operation: The user enters information about their hobbies and preferences into the input form on the device and presses the "Send" button. The device then sends this information to the server.

[1427] Step 2:

[1428] The server generates a prompt based on the received user information.

[1429] Input: User's hobbies and preferences

[1430] Output: Generated prompt sentence (e.g., "The user's hobby is fashion, and their preference is stylish cityscapes.")

[1431] Specific operation: The server analyzes the received information and generates a prompt sentence based on the template.

[1432] Step 3:

[1433] The server sends a request to the generative AI model using the generated prompt sentence.

[1434] Input: Generated prompt text

[1435] Output: Profile photo URL returned by the generative AI model

[1436] Specific operation: The server sends the generated prompt to a generative AI model (e.g., OpenAI's API) and requests the generation of a profile photo. The generative AI model generates an image based on the prompt and returns its URL to the server.

[1437] Step 4:

[1438] The server receives the profile picture URL returned by the generative AI model and downloads the image.

[1439] Input: Profile photo URL

[1440] Output: Downloaded profile photo

[1441] What happens: The server uses the returned URL to download the profile picture.

[1442] Step 5:

[1443] The server adds the downloaded profile picture to the user's profile information and sends it to the terminal.

[1444] Input: Downloaded profile photo

[1445] Output: Updated user profile information

[1446] Specific operation: The server adds the downloaded profile picture to the user's profile information and sends the updated profile information to the terminal.

[1447] Step 6:

[1448] The user uses the device to view the generated profile picture.

[1449] Input: Updated user profile information

[1450] Output: The generated profile picture displayed on the device.

[1451] Specific operation: The user checks the generated profile picture through the device, and the device displays the updated profile information received from the server.

[1452] Example 2

[1453] 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."

[1454] Conventional systems have the problem of being unable to fully utilize users' behavioral history and profile information, resulting in low accuracy in recommending optimal matches. Furthermore, insufficient data preprocessing reduces the learning efficiency of the AI ​​model, resulting in poor quality recommendation results.

[1455] 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.

[1456] In this invention, the server includes means for collecting user behavior history and profile information, means for preprocessing the collected data, means for training an AI model using the preprocessed data, and means for recommending optimal matching partners using the trained AI model. This makes it possible to effectively utilize the user's behavior history and profile information and recommend optimal matching partners with high accuracy.

[1457] "User behavior history" is a record of the operations and activities that a user performs on the system.

[1458] "Profile information" refers to personal information and attribute information provided by a user at the time of registration.

[1459] "Means of collection" refers to the methods and technologies used to store user behavior history and profile information in a database.

[1460] "Preprocessing means" refers to methods and techniques that process collected data through cleaning, normalization, feature engineering, etc., to convert it into a format that is easy for an AI model to learn.

[1461] "Means for training an AI model" refers to methods or techniques that use preprocessed data to apply machine learning algorithms to learn user behavior patterns and attributes.

[1462] "Means for recommending optimal matching partners" refers to methods and technologies that use trained AI models to predict and present optimal matching partners to users.

[1463] This invention is a system that recommends optimal matches by collecting user behavior history and profile information and generating an AI model based on this information. Specific embodiments of this system are described below.

[1464] First, the server collects the user's behavioral history and profile information. The behavioral history includes the actions the user took on the website or application (e.g., page views, message sending, likes). The profile information includes the personal information and attribute information (e.g., age, gender, hobbies, occupation) entered by the user when registering. This data is stored on a database server (e.g., MySQL, PostgreSQL). A data collection tool (e.g., Apache Kafka) is used to collect the data.

[1465] Next, the server preprocesses the collected data. Preprocessing includes data cleaning (e.g., imputing missing values ​​and removing outliers), normalization (e.g., scaling numerical data), and feature engineering (e.g., generating new features). This converts the data into a format suitable for training AI models. Data preprocessing tools (e.g., Pandas and NumPy) are used for data preprocessing.

[1466] The server then uses the preprocessed data to train an AI model using a machine learning framework (e.g., TensorFlow, PyTorch). The AI ​​model learns the user's behavioral patterns and attributes to predict the best possible match.

[1467] Finally, the server uses the trained AI model to recommend the best possible match for the user. Specifically, the user's behavioral history and profile information are input into the AI ​​model, and the model presents the best possible match to the user. The recommendation system uses recommendation algorithms (e.g., Scikit-learn, LightGBM).

[1468] Examples:

[1469] If user A has hobbies of "hiking" and "cooking" and frequently communicates with other users who have these hobbies, the AI ​​will learn this tendency. Next, if user B also has hobbies of "hiking" and "cooking," the AI ​​will recommend user A and user B as potential matches.

[1470] Example prompt sentence:

[1471] "Generate an AI model that recommends optimal matches based on a user's behavioral history and profile information. User A's hobbies are hiking and cooking, and he frequently communicates with other users who share these hobbies. User B's hobbies are also hiking and cooking. Based on this information, create an AI model that recommends User A and User B as potential matches."

[1472] The above is a specific embodiment for carrying out the present invention.

[1473] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1474] Step 1:

[1475] Data collection

[1476] The server collects the user's behavioral history and profile information.

[1477] Input: User operation history (e.g., page viewing, message sending, likes) and registration information (e.g., age, gender, hobbies, occupation)

[1478] Specific operations: The server logs the actions users take on websites and applications and stores them in a database. Data collection tools (e.g., Apache Kafka) are used to collect data in real time.

[1479] Output: User behavior history and profile information stored in a database

[1480] Step 2:

[1481] Data Preprocessing

[1482] The server pre-processes the collected data.

[1483] Input: User behavior and profile information stored in a database

[1484] Specific operations: The server performs data cleaning (e.g., missing value imputation, outlier removal), normalization (e.g., scaling numerical data), and feature engineering (e.g., generating new features) using data preprocessing tools (e.g., Pandas, NumPy).

[1485] Output: Preprocessed dataset

[1486] Step 3:

[1487] Training an AI model

[1488] The server uses the preprocessed data to train the AI ​​model.

[1489] Input: Preprocessed dataset

[1490] Specific operation: The server uses machine learning algorithms (e.g., deep learning, gradient boosting) to learn user behavior patterns and attributes. It uses machine learning frameworks (e.g., TensorFlow, PyTorch).

[1491] Output: A trained AI model

[1492] Step 4:

[1493] Matching partner recommendations

[1494] The server uses a trained AI model to recommend the best possible match for the user.

[1495] Input: Trained AI model, user behavior history and profile information

[1496] Specific operation: The server inputs the user's behavioral history and profile information into the AI ​​model, and presents the user with the best match predicted by the model. It uses a recommendation system (e.g., Scikit-learn, LightGBM).

[1497] Output: A list of best matches presented to the user

[1498] The above is the specific processing flow of the program of this system.

[1499] (Application example 2)

[1500] 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."

[1501] Conventional systems were unable to fully utilize a user's behavioral history and profile information, making it difficult to recommend optimal matches and products. Furthermore, there was a tendency for matching volume to be biased, reducing user satisfaction. Furthermore, it was difficult to make highly accurate product recommendations based on a user's interests and preferences.

[1502] 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.

[1503] In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches using advanced attribute judgments using AI, means for reducing bias in the amount of matches, means for analyzing the user's behavioral history and profile information to recommend optimal products, and means for vectorizing product descriptions and calculating cosine similarity to recommend similar products. This makes it possible to recommend optimal matches and products with high accuracy based on the user's behavioral history and profile information.

[1504] "AI-based profile picture creation method" refers to a method that uses artificial intelligence technology to generate a profile picture based on a user's characteristics and preferences.

[1505] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match and products.

[1506] "Means for inducing potential matches" refers to means for finding and recommending potential matches based on the user's attributes and behavioral history.

[1507] "Means for reducing bias in the amount of matching" refers to means for reducing bias in matching for particular users and providing equal opportunities for matching.

[1508] "Means for analyzing a user's behavioral history and profile information and recommending optimal products" refers to means for recommending optimal products to a user based on the user's past behavior and profile information.

[1509] "Means for vectorizing product descriptions, calculating cosine similarity, and recommending similar products" refers to a means for finding and recommending similar products by converting product descriptions into numerical vectors and calculating cosine similarity.

[1510] A system for implementing this invention includes a server, a user terminal, and an AI model. The server collects user behavioral history and profile information and runs software to analyze this data. Specifically, the server processes the data using libraries such as Python, Pandas, and Scikit-learn.

[1511] The server first loads the user's behavioral history data and product data. Next, it uses TfidfVectorizer to vectorize the product descriptions and extract the user's interests. This allows it to obtain vectors of products that the user has shown interest in in the past. It then calculates the cosine similarity and recommends the most similar products.

[1512] The user terminal is a device such as a smartphone or PC that displays the recommendation results sent from the server. The user can then check and purchase the recommended products.

[1513] For example, if a user frequently purchases "sports goods," the server will recommend the latest sports goods and related products based on that user's behavioral history, allowing the user to easily find products that match their interests.

[1514] An example of a prompt is as follows:

[1515] "Create a Python program that recommends optimal products based on user behavior history data. Include a function to extract user interests and calculate cosine similarity to recommend the most similar products."

[1516] In this way, the system for implementing the invention is able to recommend optimal matches and products with a high degree of accuracy based on the user's behavioral history and profile information.

[1517] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1518] Step 1:

[1519] The server reads user behavior history data and product data. Specifically, it obtains data from files named user_behavior.csv and product_catalog.csv. The input is user behavior history data and product data, and the output is a data frame containing these data. The server reads these data using the Pandas library.

[1520] Step 2:

[1521] The server vectorizes the product description. Specifically, it converts the product description into a numeric vector using TfidfVectorizer. The input is the product description text, and the output is the vectorized product description. The server performs this process using the Scikit-learn library.

[1522] Step 3:

[1523] The server obtains vectors of products of interest from the user's behavioral history. Specifically, it extracts corresponding vectors based on the IDs of products that the user has previously shown interest in. The input is the user's behavioral history data and vectorized product descriptions, and the output is a vector that reflects the user's interests.

[1524] Step 4:

[1525] The server calculates the cosine similarity between the vector reflecting the user's interests and the vector of all products. The input is the vector reflecting the user's interests and the vector of all products, and the output is the similarity score. The server performs this calculation using the Scikit-learn library.

[1526] Step 5:

[1527] The server recommends the most similar products. Specifically, it selects the product with the highest score based on the cosine similarity score and recommends it to the user. The input is the similarity score, and the output is a list of recommended products.

[1528] Step 6:

[1529] The terminal displays the recommendation results sent from the server. Specifically, it displays a list of recommended products on the user's smartphone or PC. The input is the recommendation results sent from the server, and the output is the product list displayed to the user.

[1530] Step 7:

[1531] The user checks the recommended products and purchases them. Specifically, they select products of interest from the product list displayed on the terminal and then carry out the purchase procedure. The input is the list of recommended products, and the output is information about the purchased product.

[1532] Example 3

[1533] 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."

[1534] In conventional systems, there was a problem that the amount of matching between users was easily biased, and some users did not get matching opportunities. In addition, there was an issue that appropriate recommendations were not made to users with few matches, resulting in a poor user experience.

[1535] 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.

[1536] In this invention, the server includes means for storing the matching status of users in a database, means for the server to periodically acquire the matching status of all users, means for the server to input the matching status to the AI ​​model, means for the AI ​​model to identify users with few matches, means for the AI ​​model to select recommended users for the identified users, and means for the server to notify target users of information about the recommended users. This makes it possible to reduce bias in the amount of matches and make appropriate recommendations to users with few matches.

[1537] "Means for saving a user's matching status in a database" is a function that records information about a new match made by a user in a database.

[1538] "Means for the server to periodically obtain the matching status of all users" refers to a function by which the server obtains the matching status of all users from the database at regular time intervals.

[1539] "Means for the server to input the matching status into the AI ​​model" refers to the function by which the server inputs the matching status data it has acquired into the AI ​​model.

[1540] "Means for the AI ​​model to identify users with few matches" refers to a function in which the AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time.

[1541] "Means for selecting recommended users for a user identified by an AI model" is a function that selects recommended users for a user identified by an AI model based on attributes and past matching history.

[1542] "Means for the server to notify the target user of information about recommended users" is a function by which the server notifies the target user of information about recommended users selected by the AI ​​model.

[1543] The "scheduler" is a function that allows the server to automate the process of periodically obtaining the matching status of all users.

[1544] "User attributes" refers to personal information such as the user's age, hobbies, and region.

[1545] MODE FOR CARRYING OUT THE INVENTION

[1546] The present invention relates to a system that reduces bias in the amount of matching for users and makes appropriate recommendations to users with few matches. Specific embodiments of this system will be described below.

[1547] System configuration

[1548] This system consists of three main elements: a server, a terminal, and a user. The server contains software components such as a database, an AI model, and a scheduler. A terminal is a device used by a user (such as a smartphone, tablet, or PC). A user is an individual who uses the system.

[1549] Hardware and software used

[1550] Server: Use a high-performance server machine (e.g., AWS EC2 instance). Use MySQL as the database and TensorFlow or PyTorch to build the AI ​​model.

[1551] Device: A user device such as a smartphone, tablet, or PC.

[1552] Scheduler: Use scheduling tools such as cron jobs to automate periodic data retrieval.

[1553] Program processing

[1554] The server stores the matching status of users in a database. When a user makes a new match, the device records that information in the database. The server retrieves the matching status of all users from the database at regular intervals (e.g., every hour). This process is automated using a scheduler.

[1555] The server inputs the acquired matching status data into the AI ​​model. The AI ​​model performs analysis based on the user's attributes (age, hobbies, region, etc.) and past matching history. The AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time (e.g., one week). These identified users are listed as users with few matches.

[1556] The AI ​​model recommends other users who are likely to have matching attributes and hobbies to the identified user. The recommendation algorithm is based on the user's attributes, such as age, hobbies, and region. The server notifies the target user of information about the recommended users selected by the AI ​​model. Notifications are sent to the device via push notification or email. By receiving the notification, users gain new matching opportunities.

[1557] Specific examples

[1558] For example, if User A (age 30, hobby: hiking, location: Tokyo) has not had a match for more than a week, the server retrieves User A's matching status from the database. The server inputs this data into the AI ​​model, which then recommends User B (age 28, hobby: hiking, location: Tokyo), who is likely to be a good match for User A. The server then sends User A a push notification with User B's profile. User A receives the notification and has the opportunity to newly match with User B.

[1559] Prompt Sentence Examples

[1560] "If User A, age 30, has not had a match for more than a week, the AI ​​should recommend User B, age 28, who enjoys hiking."

[1561] In this way, the system processing can be performed while clarifying the roles of the server, terminal, and user. The flow of the specific processing in the third embodiment will be described with reference to FIG.

[1562] Step 1:

[1563] When a user makes a new match, the device stores that information in a database.

[1564] Input: User's matching information (user ID, matched person's ID, matching date and time)

[1565] Output: Matching information stored in a database

[1566] Specific operation: When a user matches, the device records the user ID, the matched person's ID, and the date and time of the match in a database.

[1567] Step 2:

[1568] The server periodically obtains the matching status of all users.

[1569] Input: All user matching information stored in the database

[1570] Output: Matching status data for all users obtained by the server

[1571] Specific operation: The server uses a scheduler to retrieve the matching status of all users from the database at regular intervals (e.g., every hour).

[1572] Step 3:

[1573] The matching status data acquired by the server is input into the AI ​​model.

[1574] Input: Matching status data of all users acquired by the server

[1575] Output: Matching situation data input into the AI ​​model

[1576] Specific operation: The server converts the acquired matching status data into an appropriate format and passes it to the AI ​​model in order to input it into the AI ​​model.

[1577] Step 4:

[1578] The AI ​​model identifies users with low matches.

[1579] Input: Matching situation data input into the AI ​​model

[1580] Output: List of users with few matches

[1581] Specific operation: The AI ​​model analyzes the matching status of all users and identifies users who have not had a match for a certain period of time (e.g., one week).

[1582] Step 5:

[1583] The AI ​​model selects recommended users for the identified user.

[1584] Input: List of users with few matches, user attribute data

[1585] Output: A list of recommended users

[1586] Specific operation: The AI ​​model selects recommended users for the identified user based on their attributes and past matching history.

[1587] Step 6:

[1588] The server notifies the target user of the information about the recommended user.

[1589] Input: List of recommended users

[1590] Output: Information about the recommended user notified to the target user

[1591] Specific operation: The server notifies the target user of the information about the recommended users selected by the AI ​​model. Notifications are sent to the device via push notification or email.

[1592] (Application example 3)

[1593] 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."

[1594] Conventional systems tend to be prone to bias in the amount of matching a user receives, and there was a problem in that appropriate responses were not made to users who had not been matched for a certain period of time. Furthermore, even on online shopping sites, it was difficult to effectively recommend specific products because users' purchase and browsing histories could not be fully utilized. This led to issues such as lower user satisfaction and a decrease in the frequency of site use.

[1595] 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.

[1596] In this invention, the server includes a means for creating a profile picture using AI, a means for inducing potential matches using advanced attribute judgment by AI, a means for reducing bias in the amount of matches, a means for analyzing a user's purchase history and browsing history and recommending specific products, and a means for actively recommending products to users who have not made a purchase for a certain period of time. This makes it possible to reduce bias in the amount of matches for users and effectively recommend specific products.

[1597] A "means for creating a profile picture" is a means for using AI to generate a profile picture based on a user's characteristics and preferences.

[1598] "Advanced attribute judgment" means that AI analyzes a user's behavioral history and profile information to make judgments to recommend the most suitable match.

[1599] "Means to reduce imbalances in the amount of matches" refers to a method in which AI monitors the matching status of all users and actively recommends potential matches to users with few matches.

[1600] "Means for analyzing purchase history and browsing history" refers to a means by which AI analyzes a user's past purchase history and browsing history to understand their interests and concerns.

[1601] "Means of recommending specific products" refers to a means of recommending products that match the user based on the user's purchasing history and browsing history analyzed by AI.

[1602] The "means for actively recommending products to users who have not made a purchase for a certain period of time" refers to a means for identifying users who have not made a purchase for a certain period of time and actively recommending products to those users.

[1603] The system for implementing this invention consists of three main elements: a server, a terminal, and a user. The server includes a means for creating profile pictures using AI, a means for determining advanced attributes, a means for reducing bias in the amount of matching, a means for analyzing purchase history and browsing history, a means for recommending specific products, and a means for proactively recommending products to users who have not made a purchase for a certain period of time.

[1604] Server Processing

[1605] The server first collects user profile information and behavioral history, then analyzes this data using AI. Specifically, it uses the pandas library to load user and product data, and uses scikit-learn's cosine_similarity to recommend products based on the user's attributes. This allows it to generate a profile picture based on the user's characteristics and preferences, and recommend optimal matches and products.

[1606] Terminal handling

[1607] The device provides an interface for users to access the service. Through a device such as a smartphone or PC, users can create a profile picture and check the results of product recommendations. In addition, if a user has not made a purchase for a certain period of time, the system will actively recommend products to them via push notifications and emails.

[1608] User operations

[1609] Users enter their profile information through their devices and provide their purchase and browsing history. This allows the server to accurately understand the user's attributes and recommend the most suitable match and products. The user can then check the recommended products and purchase them.

[1610] Specific examples

[1611] For example, if a female user in her 30s is interested in fashion and has previously purchased casual wear and sneakers, the server will analyze the user's purchasing history and recommend the following products:

[1612] Example prompt sentence:

[1613] User ID: 12345

[1614] Attributes: Women in their 30s, fashion enthusiasts, mid-range price range

[1615] Purchase history: casual wear, sneakers, accessories

[1616] Recommended products:

[1617] casual jacket

[1618] Sports shoes

[1619] Simple necklace

[1620] By inputting this prompt sentence into a generative AI model, the most suitable product recommendations are made to the user.

[1621] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1622] Step 1:

[1623] The server reads user data and product data. Specifically, it reads data from a CSV file using the pandas library. The input is the CSV file of user data and product data, and the output is the user data and product data in data frame format.

[1624] Step 2:

[1625] The server analyzes the user's purchase history and browsing history. Specifically, it extracts the purchase history of a specific user and obtains the user's attribute information. The input is the user ID, and the output is the user's purchase history data.

[1626] Step 3:

[1627] The server uses an AI model to recommend products based on user attributes. Specifically, it uses scikit-learn's cosine_similarity to calculate the similarity between the user's attribute vector and the product vector. The input is the user's attribute information and product data, and the output is a list of products with high similarity.

[1628] Step 4:

[1629] The server identifies users who have not made a purchase for a certain period of time. Specifically, it compares the current date with the user's last purchase date and extracts users who have not made a purchase for a certain period of time (e.g., 30 days). The input is user data, and the output is a list of users who have not made a purchase for a certain period of time.

[1630] Step 5:

[1631] The server proactively recommends products to users who have not made a purchase for a certain period of time. Specifically, it uses the product list obtained in step 3 to send push notifications and emails to users. The input is a list of users who have not made a purchase for a certain period of time and a product list, and the output is a notification of recommended products sent to the user.

[1632] Step 6:

[1633] The terminal provides an interface that allows users to check recommended products. Specifically, it displays recommended products on the screen of a smartphone or PC. The input is recommended product data sent from the server, and the output is recommended product information that is displayed to the user.

[1634] Step 7:

[1635] The user checks the recommended products through the terminal and purchases them. Specifically, they click on the recommended products to check detailed information and then complete the purchase procedure. The input is the recommended product information displayed on the terminal, and the output is the completion of the purchase procedure.

[1636] 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.

[1637] "Example 1"

[1638] In one embodiment of the present invention, in addition to a means for creating a profile picture using AI, a means for determining advanced attributes using AI, and a means for reducing bias in the amount of matches, an emotion engine that recognizes the user's emotions is incorporated. This emotion engine recognizes emotions from the user's text input or voice input and recommends matching partners based on those emotions. For example, if a user inputs the text "I'm having fun," the emotion engine recognizes that the user is currently in a happy mood and recommends matching partners that match that mood.

[1639] "Example 2"

[1640] In another embodiment of the present invention, the emotion engine tracks changes in the user's emotions and re-recommends matches based on those changes. For example, if a user first inputs the text "I'm happy" and then later inputs the text "I'm lonely," the emotion engine recognizes that the user's emotions have changed from happy to lonely and re-recommends matches that match the user's new emotions.

[1641] "Example 3"

[1642] Furthermore, in another embodiment of the present invention, the emotion engine learns changes in the user's emotions and recommends matching partners based on the learning results. For example, if a user tends to input the text "fun" every weekend, the emotion engine learns this tendency and recommends matching partners that match the user's happy mood on weekends.

[1643] The processing flow of each embodiment will be described below.

[1644] "Example 1"

[1645] Step 1: The user enters text or voice input into the application.

[1646] Step 2: The emotion engine recognizes emotions from the user's input. For example, if the text "fun" is entered, the emotion engine recognizes that the user is in a happy mood.

[1647] Step 3: Based on the emotions recognized by the emotion engine, the AI ​​recommends potential matches, taking into account the user's profile information and past behavioral history.

[1648] "Example 2"

[1649] Step 1: The user enters text or voice input into the application.

[1650] Step 2: The emotion engine recognizes emotions from the user's input and tracks changes in those emotions. For example, if a user first inputs "happy" and then later inputs "lonely," the emotion engine will recognize the change in emotion.

[1651] Step 3: Based on the changes in emotions recognized by the emotion engine, the AI ​​re-recommends potential matches, taking into account the user's profile information and past behavioral history.

[1652] "Example 3"

[1653] Step 1: The user enters text or voice input into the application.

[1654] Step 2: The emotion engine recognizes emotions from the user's input and learns changes in those emotions. For example, if a user tends to input "fun" every weekend, the emotion engine will learn that tendency.

[1655] Step 3: Based on the emotional trends learned by the emotion engine, the AI ​​recommends potential matches, taking into account the user's profile information and past behavioral history.

[1656] Example 1

[1657] 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."

[1658] In conventional systems, it was difficult to fully reflect the user's characteristics and emotions when generating a user's profile picture or recommending potential matches. Furthermore, since matchmaking recommendations were not based on the user's emotions, there was a risk of user satisfaction decreasing. This led to a need for improved matching accuracy and user experience.

[1659] 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.

[1660] In this invention, the server includes means for inputting user profile information, means for receiving the input information, means for inputting a prompt sentence to the generative AI model, means for the generative AI model to generate a profile picture, means for analyzing the user's emotions using an emotion engine, means for recommending matching partners based on the emotions, and means for the user to confirm the generated profile picture and the recommended matching partners. This makes it possible to generate a profile picture that reflects the user's characteristics and emotions, and to recommend optimal matching partners based on the emotions.

[1661] The "means for inputting user profile information" refers to a device or software that provides an interface for users to input information such as self-introduction, hobbies, and preferences.

[1662] "Means for receiving input information" refers to a device or software that has the functionality to allow the server to receive, store, or process the profile information input by the user.

[1663] "Means for inputting prompt sentences into a generative AI model" refers to a device or software for inputting prompt sentences generated based on a user's profile information into an AI model.

[1664] A "means for a generative AI model to generate a profile picture" is a device or software that uses AI technology to generate a profile picture based on a user's characteristics and preferences.

[1665] The "means for analyzing user emotions using an emotion engine" is a device or software for analyzing emotions from a user's text input or voice input.

[1666] The "means for recommending a match based on emotion" is a device or software for recommending the most suitable match based on the analyzed emotion information of the user.

[1667] The "means for the user to review the generated profile picture and recommended matches" refers to a device or software that provides an interface that allows the user to review the generated profile picture and recommended matches.

[1668] MODE FOR CARRYING OUT THE INVENTION

[1669] This invention is a system that generates a profile picture using a generative AI model based on a user's profile information, analyzes the user's emotions using an emotion engine, and recommends optimal matches. Specific embodiments of this system are described below.

[1670] Program Generation

[1671] The server creates a program to generate a user's profile picture. This program uses a generative AI model to generate a profile picture based on information such as a self-introduction, hobbies, and preferences provided by the user. It also includes a function to analyze the user's emotions using an emotion engine and recommend potential matches based on those emotions.

[1672] Program processing

[1673] The server receives profile information entered by the user through the device. For example, the user might enter information such as "I like the outdoors and often go camping." Based on this information, the server generates a prompt sentence and inputs it into the generative AI model. Image generation models such as "Stable Diffusion" and "DALL-E" can be used as the generative AI model.

[1674] The generative AI model generates a profile picture based on a prompt received from the server. For example, the following prompt is input to the generative AI model:

[1675] The user's hobby is the outdoors, and they like camping. Based on this information, generate a profile picture with an outdoor scene in the background, such as a campsite or a mountain.

[1676] Based on this prompt, the generative AI model generates a profile photo with a campsite as the background.

[1677] Furthermore, the server uses an emotion engine to analyze the user's emotions. If the user inputs "I'm in a happy mood today," the emotion engine analyzes the information and recognizes that the user is in a happy mood. The emotion engine can use text analysis or voice analysis technology.

[1678] The server then recommends a match based on the user's emotional information. For example, if a user is in a happy mood, the server recommends users who are in a similar happy mood or who share fun activities.

[1679] Specific examples

[1680] If the user inputs "I like the outdoors and often go camping," the server will input the following prompt sentence into the generative AI model:

[1681] The user's hobby is the outdoors, and they like camping. Based on this information, generate a profile picture with an outdoor scene in the background, such as a campsite or a mountain.

[1682] Based on this prompt, the generative AI model generates a profile photo with a campsite as the background.

[1683] For example, if a user types "I'm in a happy mood today," the server will use its emotion engine to analyze the user's emotions and recommend a match that matches their happy mood. For example, it will recommend users who also like the outdoors and are in a happy mood.

[1684] In this way, the server can use AI technology based on the information provided by the user to generate profile pictures and recommend potential matches.

[1685] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1686] Program processing flow

[1687] Step 1: User enters profile information

[1688] The user inputs profile information such as a self-introduction, hobbies, and preferences through the terminal. For example, the user may input information such as "I like the outdoors and often go camping."

[1689] Input: User profile information

[1690] Output: Sending profile information from device to server

[1691] Step 2: The server receives the entered information

[1692] The server receives the profile information sent by the user from the terminal, which is used in subsequent processing.

[1693] Input: Profile information sent from the device

[1694] Output: Profile information stored on the server

[1695] Step 3: The server inputs a prompt to the generative AI model.

[1696] The server generates a prompt based on the received profile information and inputs it into the generative AI model. For example, it generates a prompt such as, "The user's hobby is the outdoors, and they like camping. Based on this information, please generate a profile picture with an outdoor scene, such as a campsite or mountain, as the background."

[1697] Input: Profile information stored on the server

[1698] Output: The prompt sentence that is input to the generative AI model

[1699] Step 4: Generative AI model generates profile photo

[1700] Generative AI models (such as "Stable Diffusion" and "DALL-E") generate profile pictures based on prompts received from the server, with backgrounds that match the user's interests and preferences.

[1701] Input: A prompt sentence entered into the generative AI model

[1702] Output: Generated profile picture

[1703] Step 5: The server uses the emotion engine to analyze the user's emotions.

[1704] The server uses an emotion engine to analyze emotions from text or voice input by the user. For example, if a user inputs "I'm in a happy mood today," the emotion engine analyzes the information and recognizes that the user is in a happy mood.

[1705] Input: User-entered text or speech

[1706] Output: Analyzed user emotion information

[1707] Step 6: The server recommends matches based on emotions

[1708] The server recommends matching partners based on the emotional information analyzed by the emotion engine. For example, if a user is in a happy mood, the server will recommend users who are in a similar happy mood or who can share fun activities.

[1709] Input: Analyzed user emotion information

[1710] Output: A list of suggested matches

[1711] Step 7: User reviews generated profile photo and suggested matches

[1712] Users can view the profile picture generated and the recommended matches through their device, and decide their next action based on this information.

[1713] Input: A generated profile picture and a list of suggested matches

[1714] Output: User confirmation and next action

[1715] (Application example 1)

[1716] 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."

[1717] In conventional systems, the generation of user profile pictures and avatars does not adequately reflect the user's characteristics and preferences, resulting in low user satisfaction. In addition, since matching does not take the user's emotions into consideration, matching accuracy is low, and the user experience is not improved.

[1718] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means. In this invention, the server includes means for creating a profile picture using AI, means for inducing potential matches through advanced attribute judgment by AI, means for reducing bias in the amount of matches, means for recognizing a user's emotions and adjusting the facial expression and pose of an avatar based on the emotions, and means for generating an avatar based on the user's characteristics and preferences. This makes it possible to generate a profile picture and avatar based on the user's characteristics and preferences, and furthermore, since matching is performed taking the user's emotions into consideration, it is possible to improve user satisfaction and matching accuracy.

[1719] "Means for creating profile pictures using AI" refers to a means for generating profile pictures using artificial intelligence technology based on information such as a self-introduction, hobbies, and preferences provided by the user.

[1720] "Advanced attribute judgment using AI" is a judgment method that uses artificial intelligence technology to analyze a user's behavioral history and profile information and recommend the most suitable match.

[1721] "Means for inducing potential matches" refers to means for finding and recommending potential matches based on the user's characteristics and preferences.

[1722] "Means for reducing bias in the amount of matching" refers to means for avoiding concentration of matching on specific users and for achieving a balance in overall matching.

[1723] "Means for recognizing a user's emotions and adjusting the avatar's facial expression and pose based on those emotions" refers to means for recognizing emotions from a user's text input or voice input and automatically changing the avatar's facial expression and pose in accordance with those emotions.

[1724] "Means for generating an avatar based on a user's characteristics and preferences" refers to a means for using artificial intelligence technology to generate an avatar that is optimal for a user based on information such as a self-introduct...

Claims

1. A creating means for creating a profile picture of the user based on the user's behavior history and profile information by inputting a specific prompt sentence into a generating AI model; an emotion engine that recognizes the emotion of the user; A means for recommending a match for the user based on the emotion using a generative AI model; means for tracking changes in the user's emotions; means for re-recommending a match partner for the user based on the change in emotion; A means for monitoring the matching status of all users and recommending a matching partner to a user whose number of matches is less than that of a specific user; A system including:

2. The system of claim 1 , further comprising means for adjusting a facial expression and a pose of the user's avatar based on the user's characteristics, the user's preferences, and the emotions.

3. The system of claim 1 , wherein the profile information includes attributes of the user and past matching history of the user.

Images