System

Abstract

To provide a system.SOLUTION: A system includes: generative AI means for transmitting a message to a user; means for inducing a user to a service based on the message transmitted by the generative AI means; means for collecting behavioral data on the user; means for analyzing the collected behavioral data and determining the intention of the user; means for generating a prompt text for instructing the generation of a message for introducing at least one of related information and other products in the same category, based on the intention of the user; means for generating a message for introducing at least one of the related information and other products in the same category, by inputting the generated prompt text into a generative AI model; and means for displaying the generated message on the terminal of the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Users have to search for much of the internet themselves. Search requires users to come up with their own words, and even the original generative AI requires users to come up with their own prompts. This is time-consuming, and it has the drawback of not being able to reach content that users cannot put into words or that they are completely unfamiliar with. [Means for solving the problem]

[0005] This invention provides a system in which generative AI sends messages to users and guides them to services. In particular, generative AI can generate messages for things that users cannot verbalize or for content they are completely unfamiliar with, and can adjust the messages based on the user's reaction. This saves users the trouble of having to search for the content themselves, and allows them to easily access unknown content. [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 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 Embodiment 1 when an emotion engine is combined. [Figure 19] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the second embodiment of the second form example when an emotion engine is combined. [Figure 20] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 of Form Example 2 when an emotion engine is combined. [Figure 21] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in the third embodiment of the third form example when an emotion engine is combined. [Figure 22] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 3 of Form Example 3 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0007] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0009] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, the processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0011] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0012] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0014] [First embodiment]

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

[0016] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0017] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0018] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0019] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0020] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0021] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0023] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0024] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0025] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

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

[0027] "Example 1"

[0028] In one embodiment of the present invention, generative AI uses natural language processing technology to understand a user's interests and preferences and generate messages based on them. The generated messages range from introducing new content the user is not aware of to responding to emotions and desires the user cannot verbalize. The generative AI learns from user responses and adjusts the message generation accordingly.

[0029] "Example 2"

[0030] As a specific example, if a generative AI determines that a user is interested in music, it will generate messages introducing new artists and songs that the user has not yet heard of. If the user responds positively to the message, the generative AI will learn from that reaction and generate similar messages more frequently. On the other hand, if the user responds negatively, the generative AI will learn from that reaction and generate similar messages less frequently.

[0031] "Example 3"

[0032] Furthermore, if generative AI determines that a user is looking for a specific product, it generates messages that introduce information related to that product or other products in the same category, thus saving users the trouble of searching for the product themselves and making it easier for them to access unknown content.

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

[0034] "Example 1"

[0035] Step 1: Generative AI uses natural language processing techniques to understand the user's interests and preferences.

[0036] Step 2: The generative AI generates a message based on the user's understood interests and preferences.

[0037] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[0038] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[0039] "Example 2"

[0040] Step 1: The generative AI determines that the user is interested in music.

[0041] Step 2: Based on that judgment, the generative AI generates a message introducing new artists and songs that the user is not yet familiar with.

[0042] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[0043] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[0044] "Example 3"

[0045] Step 1: The generative AI determines that the user is looking for a specific product.

[0046] Step 2: Based on that judgment, the generative AI generates a message introducing information related to the product and other products in the same category.

[0047] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[0048] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[0049] Example 1

[0050] 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."

[0051] Conventional generative AI systems have had difficulty fully understanding a user's interests and preferences and generating appropriate messages based on them. Furthermore, they lacked the ability to learn user responses and adjust message generation accordingly, limiting the improvement of user experience. Furthermore, it was difficult to generate messages that responded to emotions and desires that users could not verbalize. To solve these issues, a system is needed that can accurately analyze a user's interests and preferences, generate appropriate messages based on those, and learn user responses to adjust message generation.

[0052] 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.

[0053] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, a means for inputting information about the user's interests and preferences, a means for analyzing the input information, a means for generating prompts based on the analysis results, a means for generating messages based on the generated prompts, and a means for collecting user responses and feeding the collected responses back to the generative AI means. This makes it possible to accurately understand the user's interests and preferences, generate appropriate messages based on the understanding, and adjust message generation accordingly by learning from the user's responses.

[0054] "Generative AI means" refers to artificial intelligence technology for generating and sending messages to users.

[0055] "Means for guiding users to services" refers to a function for guiding users to specific services or actions based on messages sent by generative AI means.

[0056] "Means for inputting information about the user's interests and preferences" refers to an interface or function that allows the user to input information about their interests and preferences into the system.

[0057] "Means for analyzing input information" refers to a function for analyzing information input by a user using natural language processing technology, etc., and identifying the user's interests and preferences.

[0058] "Means for generating prompt sentences" refers to the function of creating sentences to instruct the generative AI means to generate messages based on the analysis results.

[0059] The "means for generating a message" refers to a function for generating an appropriate message for the user based on the generated prompt sentence.

[0060] "Means for collecting user responses and providing feedback to the generative AI means" refers to a function for collecting responses given by users to generated messages and providing that information to the generative AI means for learning.

[0061] This invention is a system that generates and provides messages to users based on their interests and preferences. The system is implemented using a server, a terminal, a generative AI model, and a natural language processing library (e.g., spaCy, NLTK).

[0062] First, a user accesses the system using a terminal and inputs information about their interests and preferences. For example, the user inputs keywords such as "travel" or "cooking." This information is then sent from the terminal to the server.

[0063] The server then analyzes the received user information using a natural language processing library. Based on the analysis results, the server identifies the user's interests and preferences. For example, if the keyword "travel" is included, the server determines that the user is interested in travel.

[0064] The server then uses the analysis results to create a prompt for the generative AI model to generate a message. For example, it generates a prompt such as, "The user seems to have recently become interested in traveling. Please generate a message that includes suggestions for new travel destinations."

[0065] The generative AI model generates a message that matches the user's interests and preferences based on the prompt it receives. For example, it might generate a message like, "Hello! I heard you've recently become interested in traveling. How about a recommended travel destination, Tuscany in Italy? You can enjoy beautiful scenery and delicious food."

[0066] The generated message is sent from the server to the user's terminal, which displays the message to the user. The user can then react to the generated message, for example, by rating the message or providing feedback. This reaction is then sent from the terminal to the server.

[0067] The server collects user responses and provides them as feedback to the generative AI model. The generative AI model learns from this feedback and reflects it in future message generation. For example, if a user rates the "Tuscany" suggestion highly, the model is more likely to suggest similar travel destinations next time.

[0068] In this way, the system understands the user's interests and preferences, and generates and delivers appropriate messages based on them. By learning the user's reactions, the system can improve the accuracy of message generation in future messages.

[0069] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0070] Step 1:

[0071] Input of user interests and preferences

[0072] Users use a terminal to input information about their interests and preferences into the system, for example, by entering keywords such as "travel" or "cooking."

[0073] Input: Keywords entered by the user about their interests and preferences.

[0074] Output: The entered keyword is sent from the terminal to the server.

[0075] Step 2:

[0076] User data analysis

[0077] The server analyzes the received information about the user's interests and preferences using a natural language processing library (e.g., spaCy, NLTK).

[0078] Input: The keyword entered by the user.

[0079] Output: Data that identifies the user's interests and preferences as a result of the analysis. For example, if the keyword "travel" is included, it is determined that the user is interested in travel.

[0080] Step 3:

[0081] Generate prompt statement

[0082] Based on the analysis results, the server creates a prompt sentence for the generative AI model to generate a message.

[0083] Input: User interests and preferences identified as a result of the analysis.

[0084] Output: A prompt to send to the generative AI model. For example, "The user seems to have recently become interested in traveling. Please generate a message with suggestions for new travel destinations."

[0085] Step 4:

[0086] Message Generation

[0087] The generative AI model generates a message that matches the user's interests and preferences based on the prompt text received.

[0088] Input: The prompt text sent by the server.

[0089] Output: A generated message to be presented to the user. For example, "Hello! I heard you've been interested in traveling lately. How about a recommended destination, Tuscany, Italy? You'll enjoy beautiful scenery and delicious food."

[0090] Step 5:

[0091] Providing a message

[0092] The server sends the generated message to the user's terminal.

[0093] Input: The generated message from the generative AI model.

[0094] Output: The message sent to the user's terminal.

[0095] Step 6:

[0096] Collecting user responses

[0097] Users can react to the generated messages, for example by rating the messages and providing feedback.

[0098] Input: User response (ratings and feedback).

[0099] Output: User responses sent from the device to the server.

[0100] Step 7:

[0101] Training generative AI models

[0102] The server collects user responses and provides them as feedback to the generative AI model.

[0103] Input: User response data.

[0104] Output: The generative AI model learns and reflects this in future message generation. For example, if a user rates the "Tuscany" suggestion highly, the model will likely suggest similar travel destinations next time.

[0105] (Application example 1)

[0106] 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."

[0107] Conventional content recommendation systems have difficulty fully understanding a user's interests and preferences and recommending appropriate content. It is also difficult to generate messages that address emotions and desires that users cannot verbalize. Furthermore, they lack the ability to continuously adjust messages based on user responses, making it difficult to improve the user experience. To solve these issues, a system is needed that can deeply understand a user's interests and preferences and recommend new content based on that understanding.

[0108] 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.

[0109] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, and a means for understanding the user's interests and preferences and recommending new content based on those. This makes it possible to recommend appropriate content based on the user's interests and preferences, and to generate messages that respond to emotions and desires that the user cannot verbalize. Furthermore, by continuously adjusting the messages based on the user's reactions, the user experience can be improved.

[0110] "Generative AI methods" are artificial intelligence methods that use natural language processing technology to understand a user's interests and preferences and generate messages based on them.

[0111] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[0112] The "means for understanding the user's interests and preferences" refers to a means for analyzing the user's past behavioral data and input information to understand the user's interests and preferences.

[0113] The "means for recommending new content" is a means for selecting and recommending new content that the user does not yet know based on the user's interests and preferences.

[0114] "Means for adjusting messages based on user responses" refers to a means by which the generative AI means learns feedback from users and reflects it in future message generation.

[0115] A system for implementing this invention includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for understanding the user's interests and preferences and recommending new content based on those interests and preferences. Specific embodiments are described below.

[0116] The server uses the OpenAI (registered trademark) API as a generative AI means. It analyzes the user's past behavioral data and input information to understand the user's interests and preferences. This analysis is performed using a programming language such as Python. To recommend new content based on the user's interests and preferences, the generative AI means generates prompts and generates messages using the OpenAI API.

[0117] As a concrete example, if the user's interests are "sci-fi," "action," and "comedy," the generated recommendation message would look like this:

[0118] "Based on your interests, we recommend the new movie 'Interstellar,' the music 'Action Movie Soundtracks,' and the article 'Latest Comedy Movie Reviews.'"

[0119] Example prompt sentence:

[0120] "The user's interests and preferences are ['sci-fi', 'action', 'comedy']. Recommend new movies, music, and articles based on this."

[0121] To tailor messages based on user responses, the generative AI method learns from user feedback and incorporates it into future message generation. This feedback is learned using machine learning algorithms.

[0122] This system makes it possible to recommend appropriate content based on the user's interests and preferences, and to generate messages that respond to emotions and desires that the user cannot verbalize. Furthermore, by continuously adjusting messages based on the user's reactions, it is possible to improve the user experience.

[0123] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0124] Step 1:

[0125] The server obtains the user's interests and preferences.

[0126] Input: User ID

[0127] Output: User's interests and preferences (e.g. "sci-fi", "action", "comedy")

[0128] Specific operation: The server retrieves interest and preference data corresponding to the user ID from the database. This data is based on the user's past behavior history and input information.

[0129] Step 2:

[0130] The server generates prompts based on the user's interests and preferences.

[0131] Input: User interests and preferences

[0132] Output: Prompt (e.g., "The user's interests and preferences are ['sci-fi', 'action', 'comedy']. Based on this, please recommend new movies, music, and articles.")

[0133] Specific operation: Based on the acquired user interests and preferences, the server generates a prompt sentence to be input into the generative AI model.

[0134] Step 3:

[0135] The server generates recommendation messages using a generative AI model.

[0136] Input: prompt statement

[0137] Output: Recommendation message (e.g., "Based on your interests, we recommend the new movie 'Interstellar,' the music 'Action Movie Soundtracks,' and the article 'Latest Comedy Movie Reviews.'")

[0138] Specific operation: The server sends the generated prompt sentence to OpenAI's API and generates a recommendation message using the generative AI model.

[0139] Step 4:

[0140] The server sends the generated recommendation message to the user.

[0141] Input: Recommendation message, user ID

[0142] Output: Recommendation message displayed on the user's device

[0143] Specific operation: The server sends the generated recommendation message to the user's terminal, allowing the user to receive the message.

[0144] Step 5:

[0145] The user sends a response to the recommendation message to the server.

[0146] Input: User response (e.g. clicks, feedback comments)

[0147] Output: User response data sent to the server

[0148] Specific operation: The user clicks or makes a feedback comment on the recommendation message, and sends the reaction data to the server.

[0149] Step 6:

[0150] The server learns the user's reaction data and reflects it in subsequent message generation.

[0151] Input: User response data

[0152] Output: An updated generative AI model

[0153] Specific operation: The server analyzes the user's reaction data and adjusts the parameters of the generation AI model to reflect the results in future message generation. This enables the server to generate messages that are more suited to the user's interests and preferences.

[0154] Example 2

[0155] 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."

[0156] Conventional generative AI systems have had difficulty accurately grasping user interests and effectively recommending new content. Furthermore, they lacked a means to appropriately provide feedback on user reactions and optimize the generative AI model, making it difficult to increase user satisfaction.

[0157] 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.

[0158] In this invention, the server includes means for collecting user behavior data, means for analyzing the collected data to determine the user's interests, a generation AI means for generating messages recommending new content based on the user's interests, means for collecting user responses to the generated messages, and means for providing feedback to the generation AI means based on the collected user responses to allow it to learn. This makes it possible to accurately grasp the user's interests and recommend appropriate content. Furthermore, optimizing the generation AI model based on the user's responses can increase user satisfaction.

[0159] "User behavior data" refers to information such as the operations and selections a user makes within the system, playback history, and ratings.

[0160] "Means for collection" refers to functions and devices for acquiring and storing user behavioral data.

[0161] "Means for analyzing and determining user interests" refers to functions and devices for analyzing collected user behavior data and identifying genres and content in which the user is interested.

[0162] "Generative AI means" refers to artificial intelligence models or algorithms that generate messages recommending new content based on a user's interests.

[0163] "Means for collecting user responses" refers to a function or device for recording and storing user responses (e.g., positive and negative responses) to generated messages.

[0164] "Means of feedback and learning" refers to functions and devices that adjust the generative AI means based on collected user responses and retrain the model.

[0165] "New content" refers to information about music, artists, videos, articles, and other things that users may not yet know about or that may be of interest to them.

[0166] "Message" refers to text or notifications generated to recommend new content to users.

[0167] MODE FOR CARRYING OUT THE INVENTION

[0168] The present invention is a system that collects and analyzes user behavior data to determine the user's interests and generates messages recommending new content. A specific embodiment of this system will be described below.

[0169] Collecting user behavior data

[0170] The server uses the APIs of music streaming services and other data sources to collect user behavior data. For example, it uses APIs of Spotify and Apple Music to obtain user playback history and rating data. This allows it to understand what kind of music users listen to and which artists they like.

[0171] User interest analysis

[0172] The server analyzes the collected behavioral data to determine the user's interests. Specifically, it uses a data analysis algorithm to identify the genres and artists the user frequently listens to. Based on the analysis results, the server understands the user's musical preferences.

[0173] Generating recommendation messages for new content

[0174] The server uses a generative AI model to generate messages that recommend new content based on the user's interests. The generative AI model receives prompt sentences that reflect the user's interests as input and generates messages based on them.

[0175] As a concrete example, the following prompt sentence is input to the generative AI model:

[0176] The genre that the user has been listening to recently is "Rock," and they especially like "Queen." Please generate a message that introduces new rock artists and songs that the user has not yet heard of.

[0177] Based on this prompt, the generative AI model generates a message like this:

[0178] "Hi! I see you've been listening to a lot of Queen lately. If you haven't heard them yet, I recommend The Black Keys. I think you'll love 'Lonely Boy'!"

[0179] Collecting user responses

[0180] The device collects user reactions to the generated messages. When a user takes an action such as "Like" or "Skip" on a message, the device records the reaction data and sends it to the server.

[0181] Feedback and training for generative AI models

[0182] The server feeds back the generated AI model based on the collected user reaction data and trains it. This allows the generated AI model to optimize recommendation messages to match the user's preferences. By increasing the frequency of messages with a high percentage of positive reactions and decreasing the frequency of messages with a high percentage of negative reactions, user satisfaction can be increased.

[0183] In this way, a system is realized that collects and analyzes user behavioral data, uses a generative AI model to generate messages recommending new content, and optimizes the model based on feedback from users.

[0184] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0185] Step 1:

[0186] The server collects user behavior data. Specifically, it uses the music streaming service's API to obtain the user's playback history and rating data. It uses the user's authentication information and API key as input and obtains the user's playback history data as output. The server stores this data in a database.

[0187] Step 2:

[0188] The server analyzes the collected data to determine the user's interests. Specifically, it uses a data analysis algorithm to identify the genres and artists the user frequently listens to. As input, it uses the playback history data stored in the database, and as output, it obtains an analysis result indicating the user's interests. The server uses this analysis result in the next step.

[0189] Step 3:

[0190] The server uses the generative AI model to generate a message recommending new content based on the user's interests. Specifically, it generates a prompt sentence that reflects the user's interests and inputs it into the generative AI model. The prompt sentence created based on the analysis results is used as input, and a recommendation message is obtained as output. The server then sends this message to the user.

[0191] Step 4:

[0192] The device collects the user's reaction to the generated message. Specifically, it records whether the user performed an action such as "like" or "skip" on the message. The device uses the user's operation information as input and obtains the user's reaction data as output. The device then transmits this reaction data to the server.

[0193] Step 5:

[0194] The server feeds back the generative AI model based on the collected user reaction data and trains it. Specifically, it classifies reactions into positive and negative and provides feedback to the generative AI model. The user reaction data is used as input, and an optimized generative AI model is obtained as output. The server uses this optimized model to generate the next message.

[0195] (Application example 2)

[0196] 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."

[0197] Conventional generative AI systems have had difficulty recommending appropriate content based on a user's interests and preferences, which has led to the issue of being unable to increase user satisfaction. Additionally, they lacked a mechanism for appropriately utilizing user responses as feedback and reflecting them in subsequent recommendations, making it difficult to continuously provide content that matches the user's interests.

[0198] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for recording user responses and adjusting the frequency of message generation based on the responses. This makes it possible to recommend content based on the user's interests and preferences, thereby increasing user satisfaction. Furthermore, by appropriately utilizing the user's responses as feedback and reflecting them in the next recommendation, it becomes possible to continuously provide content that matches the user's interests.

[0199] "Generative AI means" refers to artificial intelligence technology that generates messages based on a user's interests and preferences and sends them to the user.

[0200] "Means for guiding users to services" refers to a mechanism for guiding users to specific services or content based on messages sent by generative AI means.

[0201] "Means for recording user responses" refers to technology that records as data the responses that users show to messages sent by generative AI means.

[0202] "Means for adjusting the frequency of message generation" refers to a mechanism by which the generative AI means dynamically changes the frequency at which messages are generated based on recorded user responses.

[0203] "Content" refers to the information and entertainment that users consume, such as music, videos, and articles.

[0204] "Recommendation" refers to the act of suggesting new content based on a user's interests and preferences.

[0205] A system for implementing this invention includes a generation AI means for sending messages to users, a means for directing users to services based on the messages sent by the generation AI means, and a means for recording user responses and adjusting the frequency of message generation based on the responses.

[0206] The server uses a generative AI means to generate messages based on the user's interests and preferences and sends them to the user's device. This generative AI means is implemented using a programming language such as Python and recommends appropriate content by analyzing the user's past behavioral and reaction data.

[0207] When a user receives a message and responds positively or negatively to it, the server records the response. This record is saved in a data format such as JSON. The user's response data is used as feedback when generating the next message, and the generative AI method adjusts the frequency of message generation.

[0208] As a concrete example, consider a user using a music streaming service. A generative AI tool generates a message recommending new artists or songs that the user has not yet heard of and sends it to the user's smartphone. If the user responds positively to the message, the server records the response and increases the frequency of similar recommendations next time. On the other hand, if the user responds negatively, the server decreases the frequency of similar recommendations.

[0209] Examples of prompts include:

[0210] "When a user is recommended 'Artist A - Song 1', record the action taken if they respond positively."

[0211] In this way, the server can continually provide content based on the user's interests and preferences, increasing user satisfaction.

[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 collects users' past behavioral and reaction data, including the content they have viewed, the content they have rated, their search history, etc. The input is the user's behavioral data, and the output is a list of the collected data.

[0215] Step 2:

[0216] The server analyzes the collected data to identify the user's interests and preferences. Specifically, it uses machine learning algorithms to extract patterns of content that the user prefers. The input is a list of collected data, and the output is a profile that indicates the user's interests and preferences.

[0217] Step 3:

[0218] The server uses generative AI tools to generate messages based on the user's profile, including recommendations for new content the user does not yet know. The input is the user's profile, and the output is the generated message.

[0219] Step 4:

[0220] The server sends the generated message to the user's terminal. The input is the generated message, and the output is the message displayed on the user's terminal.

[0221] Step 5:

[0222] The user responds positively or negatively to the message displayed on the terminal. The input is the user's reaction, and the output is the reaction data.

[0223] Step 6:

[0224] The server records the user's reaction data. The input is the user's reaction data, and the output is a list of the recorded reaction data.

[0225] Step 7:

[0226] The server provides feedback based on the recorded reaction data when generating the next message. Specifically, if there are many positive reactions, it increases the number of similar recommendations, and if there are many negative reactions, it decreases the number of recommendations. The input is a list of recorded reaction data, and the output is the adjusted message generation frequency.

[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, when users wanted to find a specific product, they had to search for it themselves, which was time-consuming and labor-intensive. It was also difficult to provide appropriate information for content that users could not verbalize or were completely unfamiliar with. Furthermore, the system lacked the ability to adjust messages based on user responses, leaving a need for an improved 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 a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, a means for collecting user behavioral data, a means for analyzing the collected behavioral data to determine the user's intention, a means for generating a prompt sentence based on the user's intention, a means for inputting the generated prompt sentence into a generative AI model to generate a message, and a means for displaying the generated message on the user's device. This eliminates the need for users to search for information themselves and provides appropriate information even for content that cannot be verbalized or is completely unknown. Furthermore, adjusting the message based on the user's response can improve the user experience.

[0232] "Generative AI means" refers to means that use artificial intelligence technology to generate and send messages to users.

[0233] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[0234] The "means for collecting behavioral data" refers to a means for collecting behavioral data such as a user's search history and browsing history.

[0235] "Means for analyzing behavioral data to determine user intent" refers to means for analyzing collected behavioral data to determine what the user is looking for and what their intent is.

[0236] The "means for generating a prompt sentence" is a means for generating a prompt sentence to be input to a generative AI model based on the user's intention.

[0237] "Means for inputting into a generative AI model to generate a message" refers to means for inputting the generated prompt sentence into a generative AI model to generate a message introducing related information or other products in the same category.

[0238] The "means for displaying a message on a user's terminal" refers to a means for transmitting the generated message to the user's terminal and displaying it.

[0239] This invention relates to a system that, when it determines that a user is looking for a specific product, generates a message introducing information related to that product or other products in the same category. The system includes generative AI means, means for collecting user behavioral data, means for analyzing the behavioral data to determine the user's intention, means for generating a prompt sentence, means for inputting the prompt sentence into a generative AI model to generate a message, and means for displaying the generated message on the user's device.

[0240] The server collects the user's search history and browsing history using a database management system (e.g., MySQL (registered trademark)). The collected data is analyzed using machine learning algorithms (e.g., random forests and support vector machines) to determine whether the user is looking for a specific product. For example, if it determines that the user is looking for a "smartphone," the server generates a prompt like the following:

[0241] "User is searching for a smartphone. Generate a message that provides relevant information and introduces other smartphones in the same category."

[0242] The generated prompt is then fed into a generative AI model (e.g., OpenAI's GPT-4 (registered trademark)) to generate a message introducing related information or other products in the same category. The generated message is then sent to the user's device and displayed on a web page or application interface.

[0243] For example, if the user is searching for a "smartphone," the server generates a message like this:

[0244] "Here's information related to the phone you're looking for. We also recommend the following phones: iPhone® 13, Samsung Galaxy S21, and GOOGLE PIXEL® 6."

[0245] This system saves users the trouble of searching for information themselves and provides appropriate information for content that they cannot verbalize or are completely unfamiliar with. Furthermore, by adjusting messages based on the user's reactions, the user experience can be improved. The flow of the identification process in Example 3 will be described with reference to FIG. 15.

[0246] Step 1:

[0247] The server collects user behavior data. Specifically, it obtains information about the search keywords and products viewed by users on websites and applications from a database management system (e.g., MySQL). The input is the user's search history and browsing history, and the output is the collected behavioral data.

[0248] Step 2:

[0249] The server analyzes the collected behavioral data to determine the user's intent. Specifically, it uses machine learning algorithms (e.g., random forests and support vector machines) to analyze the user's behavioral patterns. The input is the collected behavioral data, and the output is a determination of whether the user is looking for a specific product.

[0250] Step 3:

[0251] The server generates a prompt based on the user's intent. For example, if it determines that the user is looking for a "smartphone," it generates the following prompt: "The user is looking for a smartphone. Please generate a message introducing related information and other smartphones in the same category." The input is the result of the user's intent determination, and the output is the generated prompt.

[0252] Step 4:

[0253] The server inputs the generated prompt sentence into a generative AI model to generate a message. Specifically, the prompt sentence is input into a generative AI model (e.g., OpenAI's GPT-4) to generate a message introducing related information or other products in the same category. The input is the generated prompt sentence, and the output is the generated message.

[0254] Step 5:

[0255] The server displays the generated message on the user's terminal. Specifically, it sends the generated message to the user's terminal and displays it on a web page or application interface. The input is the generated message, and the output is the message displayed on the user's terminal.

[0256] Step 6:

[0257] The user checks the displayed message and clicks to check detailed information as necessary. Specifically, when the user clicks on the displayed message, detailed information is displayed. The input is the user's click operation, and the output is the display of detailed information.

[0258] (Application example 3)

[0259] 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."

[0260] On conventional online shopping sites, when users search for a specific product, they have to search for related products or other products in the same category, which is time-consuming. It is also difficult for users to find the right product for needs they cannot verbalize or content they are completely unfamiliar with. Furthermore, there is a lack of functionality to adjust messages based on user responses, leaving a need for an improved user experience.

[0261] 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.

[0262] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for analyzing the user's search query and recommending related products or other products in the same category. This eliminates the need for the user to perform a search themselves and automatically recommends related products or other products in the same category. Furthermore, it is possible to recommend appropriate products for needs that the user cannot verbalize or content that the user is completely unfamiliar with, and by adjusting the messages based on the user's responses, the user experience can be improved.

[0263] "Generative AI methods" are artificial intelligence technologies that automatically generate related information and products based on user behavior and search queries and send them as messages.

[0264] "Means for guiding users to services" is a function for guiding users to specific services or product pages based on messages sent by generative AI means.

[0265] "Means for analyzing search queries" refers to technology that analyzes search keywords and phrases entered by users to understand their intent and relevance.

[0266] "Means for recommending related products" is a function that automatically selects and recommends products that may interest a user based on the user's search queries and behavioral history.

[0267] The "means for recommending other products in the same category" is a function for automatically selecting and recommending other products that belong to the same category as the product searched for by the user.

[0268] The system for implementing the present invention has a function of analyzing a user's search query and recommending related products or other products in the same category. A specific embodiment of this system is described below.

[0269] System Configuration

[0270] The system consists of the following main components:

[0271] 1. Generative AI methods: Artificial intelligence technology that automatically generates relevant information and products based on user behavior and search queries and sends them as messages.

[0272] 2. Means of guiding users to services: A function to guide users to specific services or product pages based on messages sent by generative AI means.

[0273] 3. Search query analysis: Techniques for analyzing search keywords and phrases entered by users to understand their intent and relevance.

[0274] 4. A means of recommending related products: A feature that automatically selects and recommends products that may be of interest to users based on their search queries and behavioral history.

[0275] 5. A means of recommending other products in the same category: A function that automatically selects and recommends other products that belong to the same category as the product searched by the user.

[0276] Hardware and software used

[0277] Hardware: Smartphones, servers

[0278] Software: Python, OpenAI API

[0279] Data processing and calculation

[0280] The server receives the search query entered by the user and inputs it as a prompt into the generative AI model. The generative AI model generates a message recommending related products or other products in the same category based on the prompt. The generated message is sent to the user's smartphone, where the user can view it.

[0281] Specific examples

[0282] If a user searches for "running shoes" on their smartphone, the server receives this search query and inputs the following prompt sentence into the generative AI model:

[0283] Example prompt sentence:

[0284] User is searching for 'running shoes'. Recommend related products or other products in the same category.

[0285] Based on this prompt, the generative AI model generates a recommendation message like the following:

[0286] An example of the message generated:

[0287] Looking for running shoes? We also recommend:

[0288] Running wear

[0289] Fitness tracker

[0290] Running socks

[0291] Sports Watch

[0292] This message is sent to the user's smartphone, allowing them to easily find related products or other products in the same category, saving the user the trouble of searching for them themselves and improving convenience.

[0293] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0294] Step 1:

[0295] The user searches for a specific product on their device (smartphone).

[0296] Input: The search query entered by the user (e.g., "running shoes")

[0297] Output: The search query is sent to the server.

[0298] Specific action: A user types "running shoes" into the search bar on their smartphone and presses the search button.

[0299] Step 2:

[0300] The server receives and parses the search query.

[0301] Input: The search query submitted by the user

[0302] Output: Search query analysis results (user intent and relevance)

[0303] Specific operation: The server receives the search query "running shoes" and analyzes it using natural language processing techniques.

[0304] Step 3:

[0305] The server generates a prompt based on the analysis results.

[0306] Input: Analysis result of search query

[0307] Output: Generated prompt statement

[0308] Specific behavior: The server generates a prompt sentence such as "The user is looking for 'running shoes'. Please recommend related products or other products in the same category."

[0309] Step 4:

[0310] The server inputs a prompt sentence into the generative AI model, which generates a message recommending related products or other products in the same category.

[0311] Input: Generated prompt statement

[0312] Output: Recommendation message

[0313] Specific operation: The server uses the OpenAI API to input a prompt sentence into the generative AI model and generate a recommendation message.

[0314] Step 5:

[0315] The server transmits the generated recommendation message to the user's terminal.

[0316] Input: Recommendation message

[0317] Output: A recommendation message that is displayed on the user's device.

[0318] Specific operation: The server sends the generated recommendation message to the user's smartphone, allowing the user to view the message.

[0319] Step 6:

[0320] The user views the recommendation messages and selects the product of interest.

[0321] Input: Recommendation message

[0322] Output: The product page selected by the user

[0323] Specific behavior: The user checks the recommendation message on their smartphone and taps on the product they are interested in to go to the detail page.

[0324] Step 7:

[0325] The server records the user's responses and accumulates the data to improve the accuracy of recommendations in future.

[0326] Input: User selected data

[0327] Output: Accumulated user behavior data

[0328] Specific operation: The server saves the user's selection data in a database and uses it to improve the accuracy of recommendations in the future.

[0329] 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.

[0330] "Example 1"

[0331] In one embodiment of the present invention, a system includes a generative AI means including an emotion engine that recognizes a user's emotion. The emotion engine infers the user's emotion from the user's facial expression, tone of voice, message context, etc. For example, if the user expresses joy, the emotion engine conveys that information to the generative AI means, which then generates a joy-related message based on that information.

[0332] "Example 2"

[0333] Another embodiment of the present invention is a system in which an emotion engine generates a message based on a user's emotion. Specifically, when a user feels sad, the emotion engine transmits that information to a generative AI means, which then generates a comforting message based on that information.

[0334] "Example 3"

[0335] Furthermore, one embodiment of the present invention is a system in which an emotion engine adjusts messages in response to changes in the user's emotions. For example, if the user initially expresses joy but suddenly expresses sadness, the emotion engine conveys this information to the generative AI means, which then adjusts the message based on this information and generates a comforting message.

[0336] The processing flow of each embodiment will be described below.

[0337] "Example 1"

[0338] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[0339] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[0340] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotions.

[0341] "Example 2"

[0342] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[0343] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[0344] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotions. In particular, if the user feels sad, the generative AI means generates a comforting message.

[0345] "Example 3"

[0346] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[0347] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[0348] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotion. In particular, if the user's emotion changes, the generative AI means adjusts the message accordingly.

[0349] Example 1

[0350] 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."

[0351] Conventional generative AI systems have had difficulty accurately understanding users' interests and preferences and generating appropriate messages based on them. Furthermore, they lacked the ability to capture users' reactions in real time and adjust messages based on their emotions, making it difficult to improve the user experience.

[0352] The identification process by the identification processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes generative artificial intelligence means, means for guiding a user to a service based on a message transmitted by the generative artificial intelligence means, means for inputting information on the user's interests and preferences, means for analyzing the input information, means for capturing the user's reactions, means for analyzing the captured reactions and recognizing the user's emotions, and means for adjusting the message based on the recognized emotions. This makes it possible to accurately understand the user's interests and preferences, capture the user's reactions in real time to recognize their emotions, and adjust the message based on the emotions.

[0353] "Generative AI means" refers to AI technology that generates messages based on a user's interests and preferences.

[0354] "Means for guiding users to a service" refers to a function for guiding users to a specific service or action based on a message sent by the generative artificial intelligence means.

[0355] "Means for inputting information about user interests and preferences" refers to an interface or method by which a user provides information about their interests and preferences to the system.

[0356] "Means for analyzing input information" refers to the technology or algorithms used to analyze the information entered by the user and understand its content.

[0357] "Means for capturing user responses" refers to technologies or devices that record a user's response (e.g., facial expression, tone of voice, text) to a generated message.

[0358] "Means for recognizing a user's emotions by analyzing captured reactions" refers to technology or algorithms for analyzing a user's captured reactions and inferring the user's emotions from those reactions.

[0359] "Means for adjusting a message based on recognized emotions" refers to a function for appropriately changing or adjusting a message generated by a generative artificial intelligence means based on the recognized emotions of a user.

[0360] MODE FOR CARRYING OUT THE INVENTION

[0361] The present invention is a system that uses generative artificial intelligence to understand a user's interests and preferences and generate messages based on them. The system includes the ability to capture the user's reactions, recognize their emotions, and adjust the messages accordingly.

[0362] Hardware and software used

[0363] Hardware: Servers, devices (PCs, smartphones, etc.)

[0364] Software: Generative AI models (e.g., GPT-4), emotion engines, natural language processing libraries (e.g., spaCy, NLTK)

[0365] System Overview

[0366] 1. Input of user interests and preferences

[0367] User: A user uses a terminal to access the system and enter information about their interests and preferences, for example, "Where would you like to go on your next vacation?"

[0368] 2. Analysis of input data

[0369] Terminal: Sends the information entered by the user to the server.

[0370] Server: The server analyzes the received data using natural language processing libraries. Specifically, it extracts keywords and context from the input text and understands the user's interests and preferences.

[0371] 3. Message Creation

[0372] Server: Based on the analysis results, the server sends a prompt to the generative AI model. An example of a prompt is, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[0373] Generative AI: Generative AI receives a prompt and generates an appropriate message, such as, "Hawaii is a recommended destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[0374] 4. Capturing User Reactions

[0375] Device: The user receives the generated message. The device captures the user's reactions (facial expressions, tone of voice, message context, etc.). For example, if the user shows a happy expression after reading the message, the device captures that expression with a camera.

[0376] 5. Emotional Recognition

[0377] Terminal: Sends captured data to the server.

[0378] Server: The server uses an emotion engine to analyze the captured data and infer the user's emotions, for example, detecting the emotion of joy from the user's facial expression.

[0379] 6. Tailor your message

[0380] Server: Based on the information from the emotion engine, sends a new prompt to the generative AI. An example of a prompt is, "The user is expressing joy. Please provide additional information related to that emotion."

[0381] Generative AI: Generative AI receives new prompts and generates additional messages, such as "I'll also introduce some recommended tourist spots in Hawaii!"

[0382] Specific examples

[0383] User: A user interested in traveling comes into the system and types, "Where would you like to go on your next vacation?"

[0384] Terminal: Sends user input to the server.

[0385] Server: Using a natural language processing library, extract the keyword "travel destination" and understand the user's interest as "travel."

[0386] Server: Sends the generative AI a prompt saying, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[0387] Generative AI: Generates a message like, "Hawaii is a great destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[0388] Device: The user receives this message and shows a look of delight.

[0389] Device: The camera captures the user's facial expressions and sends the data to the server.

[0390] Server: Uses an emotion engine to detect the emotion of joy from the user's facial expression.

[0391] Server: Sends the prompt to the generative AI: "The user is expressing joy. Please provide additional information related to that emotion."

[0392] Generative AI: Generates an additional message such as, "We'll also introduce some recommended tourist spots in Hawaii!"

[0393] The above is an embodiment of the present invention.

[0394] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0395] Step 1:

[0396] User: A user uses a terminal to access the system and enter information about their interests and preferences, for example, "Where would you like to go on your next vacation?"

[0397] Input: Text information about the user's interests and preferences

[0398] Output: User input data

[0399] Step 2:

[0400] Terminal: Sends the information entered by the user to the server.

[0401] Input: User-entered data

[0402] Output: User input data sent to the server

[0403] Step 3:

[0404] Server: The server analyzes the received data using a natural language processing library (e.g., spaCy, NLTK). Specifically, it extracts keywords and context from the input text and understands the user's interests and preferences.

[0405] Input: User-entered data

[0406] Data processing: Text analysis using natural language processing libraries

[0407] Output: Analysis of user interests and preferences

[0408] Step 4:

[0409] Server: Based on the analysis results, the server sends a prompt to a generative AI model (e.g., GPT-4). An example of a prompt is, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[0410] Input: Analysis results about user interests and preferences

[0411] Data processing: Prompt generation

[0412] Output: Prompt for the generative AI model

[0413] Step 5:

[0414] Generative AI: Generative AI receives a prompt and generates an appropriate message, such as, "Hawaii is a recommended destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[0415] Input: prompt statement

[0416] Data processing: Message generation

[0417] Output: The generated message

[0418] Step 6:

[0419] Device: The user receives the generated message. The device captures the user's reactions (facial expressions, tone of voice, message context, etc.). For example, if the user shows a happy expression after reading the message, the device captures that expression with a camera.

[0420] Input: The generated message

[0421] Output: User response data

[0422] Step 7:

[0423] Terminal: Sends captured data to the server.

[0424] Input: User response data

[0425] Output: User response data sent to the server

[0426] Step 8:

[0427] Server: The server uses an emotion engine to analyze the captured data and infer the user's emotions, for example, detecting the emotion of joy from the user's facial expression.

[0428] Input: User response data

[0429] Data processing: Emotion analysis using an emotion engine

[0430] Output: Analysis results on user sentiment

[0431] Step 9:

[0432] Server: Based on the information from the emotion engine, sends a new prompt to the generative AI. An example of a prompt is, "The user is expressing joy. Please provide additional information related to that emotion."

[0433] Input: Analysis results on user sentiment

[0434] Data processing: generating new prompts

[0435] Output: A new prompt for the generative AI model

[0436] Step 10:

[0437] Generative AI: Generative AI receives new prompts and generates additional messages, such as "I'll also introduce some recommended tourist spots in Hawaii!"

[0438] Input: new prompt text

[0439] Data processing: generating additional messages

[0440] Output: Additional messages generated

[0441] (Application example 1)

[0442] 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."

[0443] Conventional content delivery services recommend content based on users' interests and preferences, but they lack personalized recommendations that take into account the user's current emotions. Furthermore, it is difficult to generate messages that respond to emotions and desires that users cannot verbalize. This limits the improvement of the user experience.

[0444] 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: a generative AI means for sending a message to a user; a means for guiding the user to a service based on the message sent by the generative AI means; an emotion recognition means for recognizing emotions by analyzing the user's facial expressions and tone of voice; a preference understanding means for understanding the user's interests and preferences by analyzing the user's past viewing history and search history; and a content recommendation means for recommending personalized content based on the emotion recognition means and preference understanding means. This enables personalized content recommendation based on the user's current emotions and past preferences, improving the user experience.

[0445] "Generative AI methods" are artificial intelligence methods that understand a user's interests and preferences and generate messages based on them.

[0446] "Means for guiding users to services" refers to means for guiding users to specific services or actions based on messages sent by generative AI means.

[0447] The "emotion recognition means" is a means for recognizing the user's emotions by analyzing the user's facial expressions and tone of voice.

[0448] The "preference understanding means" is a means for understanding the user's interests and preferences by analyzing the user's past viewing history and search history.

[0449] The "content recommendation means" is a means for recommending personalized content to a user based on the emotion recognition means and the preference understanding means.

[0450] A system for implementing the present invention includes the following means for recommending personalized content based on a user's emotions and preferences.

[0451] System Configuration

[0452] 1. Generative AI means:

[0453] The server uses generative AI to understand the user's interests and preferences, and generates messages based on them. The generated messages address feelings and desires that the user cannot verbalize, or introduce new content that the user is completely unfamiliar with.

[0454] 2. Means of directing you to the service:

[0455] The server guides the user to specific services or actions based on the messages sent by the generative AI means, allowing the user to access the right content or services at the right time.

[0456] 3. Emotion recognition means:

[0457] The device uses a camera and microphone to recognize emotions by analyzing the user's facial expressions and tone of voice. The emotion recognition engine analyzes this data in real time to estimate the user's current emotion.

[0458] 4. Means of understanding preferences:

[0459] The server analyzes the user's past viewing and search history to understand their interests and preferences, and uses natural language processing technology to extract these from the text entered by the user.

[0460] 5. Content recommendation methods:

[0461] The server recommends personalized content to users based on emotion recognition and preference understanding. The recommendation algorithm learns user responses and is continually adjusted.

[0462] Hardware and software used

[0463] Hardware:

[0464] Devices such as smartphones and tablets

[0465] Camera and microphone

[0466] software:

[0467] Emotion recognition engines (e.g., the sentiment-analysis pipeline in the transformers library)

[0468] Natural language processing technology (e.g., TfidfVectorizer)

[0469] Similarity calculation algorithm (e.g., cosine_similarity)

[0470] Specific examples

[0471] When a user smiles into the smartphone camera, the emotion recognition engine recognizes "joy," and the server recommends action-packed thrillers or heartwarming romantic comedies based on the action movies and comedy shows the user has watched in the past.

[0472] Prompt Sentence Examples

[0473] "Analyze the user's facial expressions and tone of voice to recognize their current emotions. Understand their interests and preferences based on their past viewing and search history, and recommend the most appropriate content."

[0474] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0475] Step 1:

[0476] The device captures the user's facial expressions and tone of voice using a camera and microphone. As input, it receives image data from the camera and audio data from the microphone. These data are sent to an emotion recognition engine to estimate the user's current emotion. As output, it obtains the user's emotion (e.g., joy, sadness, surprise, etc.).

[0477] Step 2:

[0478] The server retrieves the user's past viewing and search history from a database. It uses the user ID as input to search for related history data. It then uses natural language processing technology (e.g., TfidfVectorizer) to vectorize this history data and extract the user's interests and preferences. As an output, it obtains vector data representing the user's interests and preferences.

[0479] Step 3:

[0480] The server integrates the data obtained from the emotion recognition means and the preference understanding means. As input, it receives the user's emotion data and preference vector data. Based on this data, the generative AI means generates the optimal message for the user. As output, a personalized message is obtained.

[0481] Step 4:

[0482] The server sends the generated message to the user. As input, it receives message data from the generative AI means. The message is sent to the user's device in the form of a push notification, email, etc. As output, the user receives the message.

[0483] Step 5:

[0484] Users access services based on the messages they receive. As input, they review the message content and execute the provided links or actions, which direct the user to specific content or services. As output, they obtain data on the user's actions.

[0485] Step 6:

[0486] The server collects user reaction data and feeds it back to the generative AI means. As input, it receives user action data. Based on this data, the generative AI means adjusts the message generation algorithm. As output, it obtains an improved message generation model.

[0487] Example 2

[0488] 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."

[0489] Conventional message generation systems have difficulty generating personalized messages based on users' interests and emotions, making it difficult to increase user satisfaction. Furthermore, they lacked a mechanism for learning user responses and improving the accuracy of message generation, making it difficult to provide services that meet user needs. Furthermore, they were unable to generate appropriate messages based on users' emotions, making it impossible to respond in a way that was sensitive to the user's feelings.

[0490] 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.

[0491] In this invention, the server includes a generative artificial intelligence means for sending messages to users, a means for directing users to services based on the messages sent by the generative artificial intelligence means, a means for analyzing the user's past behavioral data to determine their interests, a means for collecting user responses and adjusting the frequency of message generation based on the responses, and a means for analyzing the user's emotions and generating messages based on the emotions. This enables the generation of personalized messages based on the user's interests and emotions, thereby increasing user satisfaction. Furthermore, the server can learn user responses to improve the accuracy of message generation, enabling the provision of services that meet the user's needs. Furthermore, by generating appropriate messages based on the user's emotions, it is possible to respond in accordance with the user's emotions.

[0492] "Generative AI means" refers to means that uses artificial intelligence technology to generate and send messages to users.

[0493] The "means for guiding a user to a service" is a means for guiding a user to a specific service or action based on a message sent by the generative artificial intelligence means.

[0494] The "means for determining interests by analyzing past behavioral data" is a means for analyzing the user's past behavioral data (for example, music playback history and search history) and identifying the user's interests.

[0495] The "means for collecting user responses and adjusting the frequency of message generation based on the responses" refers to a means for collecting user responses to generated messages and adjusting the frequency of future message generation based on the responses.

[0496] The "means for analyzing the user's emotions and generating a message based on those emotions" refers to a means for analyzing the user's emotions and generating an appropriate message in accordance with those emotions.

[0497] This invention is a system that generates personalized messages based on a user's interests and emotions, and improves the accuracy of message generation by learning the user's reactions. This system includes generative artificial intelligence means, means for directing users to services, means for analyzing past behavioral data to determine interests, means for collecting user reactions and adjusting the frequency of message generation based on the reactions, and means for analyzing the user's emotions and generating messages based on the emotions.

[0498] The server collects the user's past behavioral data and analyzes it using machine learning libraries such as TENSORFLOW (registered trademark) and PyTorch. This determines the genre of music the user is interested in. For example, it retrieves the user's music history from a database over the past month and identifies the music genre the user listens to most frequently.

[0499] The server then uses a generative AI model (e.g., GPT-3 (registered trademark)) to generate a message introducing new artists and songs that the user might be interested in. The generated message is customized based on the user's interests. For example, if it determines that the user is interested in rock music, it sends the generative AI model a prompt saying, "Generate a message introducing a new rock artist." The generative AI model generates the message, "Check out the new rock band, 'The New Rockers'!"

[0500] The device displays the generated message to the user. If the user responds positively or negatively to the message, the device collects the response and sends it to the server. For example, if the device displays the generated message on the user's screen and the user clicks the "Like!" button, the device sends that information to the server. Conversely, if the user clicks the "Not Interested" button, the device also sends the information.

[0501] The server receives user responses and stores them as learning data. It increases the frequency of messages with a large number of positive responses and decreases the frequency of messages with a large number of negative responses. This learning process uses a reinforcement learning algorithm. For example, the server stores messages to which users have responded with a "Like!" in a learning database, and refers to this data when generating similar messages from the next time onwards to adjust the frequency of their generation.

[0502] Furthermore, the emotion engine analyzes the user's emotions. If the emotion engine determines that the user is sad, it passes that information on to the generative AI model. The generative AI model uses that information to generate a message to comfort the user. For example, the emotion engine analyzes the user's facial expressions and input text and determines that the user is sad. The emotion engine sends that information to the server, and the server prompts the generative AI model to "generate a message to comfort the user." The generative AI model generates a message saying, "You've had a tough day today. Why don't you listen to some relaxing music?"

[0503] Example prompt sentence:

[0504] Analyze the user's recent music history and generate messages introducing new artists or songs the user hasn't heard of. If the user responds positively, learn from that reaction and generate similar messages more frequently. If the user responds negatively, learn from that reaction and generate similar messages less frequently.

[0505] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0506] Step 1:

[0507] The server collects data on the user's past behavior. Specifically, it retrieves data from a database, such as the music history the user has played, search history, and links they have clicked. This data is used as input for analysis to identify the user's interests. Machine learning libraries such as TensorFlow and PyTorch are used for the analysis to identify the music genres the user listens to most frequently. The output is data indicating the user's interests.

[0508] Step 2:

[0509] The server uses a generative AI model (e.g., GPT-3) to generate messages introducing new artists and songs that the user might be interested in. As input, it uses the user interest data obtained in step 1. Specifically, it sends the generative AI model a prompt: "Generate a message introducing a new rock music artist." The output is the generated message. For example, the message generated is "Check out the new rock band, 'The New Rockers'!"

[0510] Step 3:

[0511] The device displays the generated message to the user. It uses the message generated in step 2 as input. Specifically, the device displays the message on the screen, and the user responds positively or negatively to the message. The output is the user's response data. For example, if the user clicks the "Like" button, that information is collected by the device.

[0512] Step 4:

[0513] The device sends the user's reaction data to the server. The user's reaction data collected in step 3 is used as input. Specifically, information on whether the user clicked the "Like" button or the "Not Interested" button is sent to the server. The output is the user's reaction data sent to the server.

[0514] Step 5:

[0515] The server receives the user's reaction data and stores it as training data. It uses the user's reaction data sent in step 4 as input. Specifically, it uses a reinforcement learning algorithm to increase the frequency of messages with many positive reactions and decrease the frequency of messages with many negative reactions. The output is updated message generation frequency data.

[0516] Step 6:

[0517] The emotion engine analyzes the user's emotions. It uses the user's facial expressions and input text as input. Specifically, it uses emotion recognition software such as OpenCV or Emotion API to determine whether the user is sad. The output is the user's emotion data.

[0518] Step 7:

[0519] The server receives the user's emotion data obtained from the emotion engine and transmits it to the generative AI model. The user's emotion data obtained in step 6 is used as input. Specifically, it sends a prompt to the generative AI model saying, "Please generate a message to comfort the user." The output is the generated comforting message. For example, a message might be generated saying, "You've had a tough day today. Why don't you listen to some relaxing music?"

[0520] (Application example 2)

[0521] 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."

[0522] Conventional content delivery services have difficulty generating personalized messages based on users' interests and emotions, which has led to insufficient improvements in user experience. Furthermore, they lack the ability to learn users' reactions and adjust the frequency of message generation, making it difficult to recommend appropriate content to users.

[0523] 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.

[0524] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, an emotion engine means for analyzing user emotions, a means for generating messages based on emotion data acquired by the emotion engine means, and a means for learning user reactions and adjusting the frequency of message generation. This enables the generation of personalized messages based on the user's interests and emotions, improving the user experience and realizing appropriate content recommendations.

[0525] "Generative AI means" refers to artificial intelligence technology for generating and sending messages to users.

[0526] "Means for guiding users to services" refers to a function for guiding users to specific services or content based on messages sent by generative AI means.

[0527] "Emotion engine means" refers to a technology for analyzing a user's emotions and acquiring the emotion data.

[0528] The "means for generating a message based on emotion data" refers to a function for generating an appropriate message for the user based on emotion data acquired by the emotion engine means.

[0529] "Means for learning user reactions and adjusting the frequency of message generation" refers to a function for analyzing user reactions and increasing or decreasing the frequency of message generation based on the results.

[0530] A system for implementing the present invention generates personalized messages for users and recommends appropriate content based on the user's emotions. A specific embodiment of this system is described below.

[0531] System Configuration

[0532] 1. Generative AI means:

[0533] The server generates and sends messages to users using a generative AI means that generates messages recommending new artists and songs based on the user's interests and past behavioral data.

[0534] 2. Means of directing you to the service:

[0535] The server directs the user to specific services and content based on messages sent by the generative AI means, which introduces new music and artists that may interest the user and directs the user to music streaming services.

[0536] 3. Emotion engine means:

[0537] The server analyzes the user's emotions using an emotion engine means that analyzes the user's text data and voice data and classifies the user's emotional state into "positive," "neutral," or "negative."

[0538] 4. Means of generating messages based on emotion data:

[0539] The server generates an appropriate message for the user based on the emotion data acquired by the emotion engine means. For example, if the user is expressing negative emotions, the server generates a message of comfort or encouragement.

[0540] 5. Learn user reactions and adjust message generation frequency:

[0541] The server learns user responses and adjusts the frequency of message generation based on the results: if there are a lot of positive responses, it generates similar messages more frequently, and if there are a lot of negative responses, it generates them less frequently.

[0542] Hardware and software used

[0543] Hardware: Servers, user devices (smartphones, tablets, PCs, etc.)

[0544] Software: Generative AI models (e.g., GPT-3), sentiment analysis libraries (e.g., TextBlob), music recommendation APIs (e.g., Spotify API)

[0545] Specific examples of processing

[0546] 1. User sentiment analysis:

[0547] The server analyzes the user's text data using an emotion engine means and classifies the user's emotions into "positive," "neutral," and "negative." For example, the server uses the TextBlob library to perform the emotion analysis.

[0548] 2. Message Creation:

[0549] The server uses generative AI tools to generate messages based on the emotion data. For example, if the user is expressing negative emotions, the server generates a message like, "Feeling down? Maybe these tracks can cheer you up."

[0550] 3. Music Recommendations:

[0551] The server uses a music recommendation API to recommend new artists and songs that might suit the user, for example, using the Spotify API to make recommendations based on the user's listening history.

[0552] Prompt Sentence Examples

[0553] I am developing an application that recommends music based on the user's emotions. Assuming that the user's emotions can be classified as positive, neutral, or negative, please generate Python code to provide messages and music recommendations according to each emotion. I will use the TextBlob library for sentiment analysis and an external music API for music recommendation.

[0554] In this way, personalized messages and appropriate content recommendations based on the user's interests and emotions can be generated.

[0555] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0556] Step 1:

[0557] The server acquires the user's text data. The input is the text data entered by the user, and the output is the text data. Specifically, the server acquires comments and feedback entered by the user into the application.

[0558] Step 2:

[0559] The server uses an emotion engine to analyze the acquired text data and classify the user's emotion. The input is the text data acquired in step 1, and the output is an emotion classification of "positive," "neutral," or "negative." Specifically, the server uses the TextBlob library to perform emotion analysis of the text.

[0560] Step 3:

[0561] The server generates a message using a generative AI means based on the emotion data acquired by the emotion engine means. The input is the emotion classification obtained in step 2, and the output is the message to be sent to the user. Specifically, a generative AI model (e.g., GPT-3) is used to generate a message according to the emotion.

[0562] Step 4:

[0563] The server sends the generated message to the user. The input is the message generated in step 3, and the output is the message displayed on the user's device. Specifically, the message is notified to the user's smartphone or PC.

[0564] Step 5:

[0565] The server receives user responses. The input is the user's response to the message (e.g., click, comment, feedback), and the output is the response data. Specifically, it records what actions the user took on the message.

[0566] Step 6:

[0567] The server analyzes the acquired user reaction data and uses it as learning data for the generative AI means. The input is the reaction data acquired in step 5, and the output is the learning result. Specifically, if there are many positive reactions, it increases the frequency of generating similar messages, and if there are many negative reactions, it decreases the frequency.

[0568] Step 7:

[0569] The server uses a music recommendation API to recommend new artists and songs suitable for the user. The input is the user's playback history and interest data, and the output is recommended music content. Specifically, it uses the Spotify API to recommend music based on the user's preferences.

[0570] Step 8:

[0571] The server provides the recommended music content to the user. The input is the music content recommended in step 7, and the output is a music recommendation list that is displayed on the user's device. Specifically, the music recommendation list is displayed on the user's smartphone or PC.

[0572] In this way, personalized messages and appropriate content recommendations based on the user's interests and emotions can be generated.

[0573] Example 3

[0574] 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."

[0575] With conventional systems, it was very time-consuming for users to search for specific products, and it was difficult to provide appropriate messages that responded to changes in the user's emotions. Furthermore, information was insufficient for content that the user could not verbalize or was completely unfamiliar with. This resulted in a poor user experience and hindered the promotion of service usage.

[0576] 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.

[0577] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, a means for collecting user behavioral data, a means for analyzing the collected behavioral data, a means for generating related information based on the analysis results, a means for detecting changes in the user's emotions, and a means for adjusting messages based on the detected changes in emotions. This eliminates the need for users to search for specific products and enables the provision of appropriate messages in response to changes in the user's emotions. It also enables the provision of information about content that the user cannot verbalize or is completely unfamiliar with.

[0578] "Generative AI means" refers to means that use artificial intelligence technology to generate and send messages to users.

[0579] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[0580] The "means for collecting behavioral data" refers to a means for collecting behavioral data such as a user's search history and click history.

[0581] The "means for analyzing behavioral data" refers to a means for analyzing collected user behavioral data and identifying the user's intentions and interests.

[0582] The "means for generating related information" is a means for generating information related to the product the user is looking for or other products in the same category based on the analysis results of the behavioral data.

[0583] The "means for detecting changes in emotions" is a means for monitoring the user's inputs and actions in real time and detecting changes in the user's emotions.

[0584] The "means for adjusting the message" refers to a means for adjusting the message using generative AI means based on detected changes in the user's emotions and changing the content to appropriate content.

[0585] This invention is a system that saves users the trouble of searching for specific products and provides appropriate messages that respond to changes in the user's emotions.It also provides information about things that the user cannot put into words or about content that the user is completely unfamiliar with.

[0586] Hardware and software used

[0587] Hardware: Servers, user devices (PCs, smartphones, etc.)

[0588] Software: Generative AI models (e.g., GPT-4), emotion engines, database management systems (e.g., MySQL)

[0589] System configuration

[0590] 1. The server collects user behavior data by storing the user's search history and click history on the device in a database.

[0591] 2. The server uses machine learning algorithms to analyze the collected behavioral data, thereby identifying the products and interests of the user.

[0592] 3. The server uses a generative AI model (e.g., GPT-4) to generate messages that provide information related to the product the user is looking for and introduce other products in the same category.

[0593] 4. The emotion engine monitors user input and behavior in real time to detect changes in the user's emotions.

[0594] 5. The server uses the generative AI model to adjust the message based on the emotional change information received from the emotion engine and sends it to the user.

[0595] Specific examples

[0596] When a user is looking for a specific product

[0597] A user searches for "smartphone."

[0598] The server collects and analyzes the search history.

[0599] The server uses the generative AI model to generate and send messages to users containing relevant information, such as "latest smartphone reviews" or "recommended smartphone accessories."

[0600] Message adjustment according to user's emotional changes

[0601] The user types, "I'm so happy I bought a new smartphone."

[0602] The emotion engine detects the emotion.

[0603] The user types, "I'm sad my phone is broken."

[0604] The emotion engine detects the change in emotion and communicates it to the server.

[0605] The server uses the generative AI model to generate a comforting message to the user, such as "I see you're having trouble. Here's how to contact our support center," and sends it to the user.

[0606] Prompt Sentence Examples

[0607] Generate product-related information

[0608] “If a user is looking for a smartphone, generate a message with relevant information and other products in the same category.”

[0609] Adapting messages to reflect changing emotions

[0610] "If the user's emotion changes from happy to sad, generate a comforting message."

[0611] In this way, it is possible to save the user the trouble of searching for a specific product and to provide appropriate messages according to changes in the user's emotions. It is also possible to provide information about content that the user cannot verbalize or is completely unfamiliar with. The flow of the identification process in the third embodiment will be described with reference to FIG. 21.

[0612] Step 1:

[0613] Collecting user behavior data

[0614] Input: Search and click history of the user on their device

[0615] What it does: When a user searches for "smartphone," the search query is sent to the server, which stores this data in a database.

[0616] Output: Stored user behavior data

[0617] Step 2:

[0618] User behavior analysis

[0619] Input: Stored user behavior data

[0620] How it works: The server uses machine learning algorithms to analyze user behavior data. For example, it determines that a user is searching for a "smartphone" based on their past search and click history.

[0621] Output: Analysis of user search results and interests

[0622] Step 3:

[0623] Generating related information

[0624] Input: Analytics about the products and interests of users

[0625] Specific operation: The server inputs a prompt to the generative AI model (e.g., GPT-4) saying, "If the user is looking for a smartphone, please generate a message that introduces related information and other products in the same category." The generative AI model then generates a message based on this.

[0626] Output: Messages generated containing relevant information

[0627] Step 4:

[0628] Detecting changes in emotions

[0629] Input: User input or actions

[0630] How it works: The emotion engine monitors user input and behavior in real time. For example, if a user types "I'm happy because I bought a new smartphone" and then "I'm sad because my smartphone broke," the emotion engine will detect this change.

[0631] Output: Information on detected changes in user emotions

[0632] Step 5:

[0633] Message coordination and generation

[0634] Input: Detected changes in user emotion

[0635] Specific operation: The server inputs a prompt to the generative AI model: "When the user's emotion changes from joy to sadness, please generate a comforting message." The generative AI model then generates a message based on this.

[0636] Output: The adjusted message

[0637] The above is the specific flow of the program processing of this system.

[0638] (Application example 3)

[0639] 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."

[0640] On traditional online shopping sites, users often spend a lot of time and effort searching for specific products. Furthermore, the lack of appropriate product suggestions and messaging tailored to the user's emotions hinders the user experience. Furthermore, the lack of suggestions for needs that users cannot verbalize or for content they are unfamiliar with limits users' opportunities to discover new products and services.

[0641] 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.

[0642] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, an emotion engine means for analyzing the user's emotions, and a means for adjusting the messages based on the emotion information analyzed by the emotion engine means. This eliminates the need for users to search for specific products, enables appropriate product suggestions based on emotions, and improves the user experience. It can also make suggestions for needs that users cannot verbalize or for unknown content.

[0643] "Generative AI means" refers to artificial intelligence technology that generates and sends messages to users.

[0644] "Means for guiding users to services" refers to technology for guiding users to specific services or product pages based on messages sent by generative AI means.

[0645] "Emotion engine means" refers to a technology for analyzing a user's emotions and acquiring that emotion information.

[0646] The "means for adjusting a message" refers to a technology for appropriately changing the message generated by the generative AI means based on the emotional information analyzed by the emotion engine means.

[0647] A system for carrying out the present invention includes a generative AI means for analyzing a user's emotions and generating an appropriate message based on the emotions, a means for directing the user to a service based on the generated message, an emotion engine means for analyzing the user's emotions, and a means for adjusting the message based on the emotion information.

[0648] System Program

[0649] The system program performs the following processing.

[0650] Hardware

[0651] Smartphone

[0652] software

[0653] Python

[0654] Sentiment analysis libraries (e.g. textblob, nltk)

[0655] Generative AI models (e.g., GPT-3)

[0656] Data processing and calculation

[0657] 1. Sentiment analysis: The sentiment analysis library analyzes the user's input text to identify the emotion. For example, if a user types "I want this smartphone," the sentiment analysis library analyzes the text and identifies the user's emotion as "happy."

[0658] 2. Message generation: The emotion engine means conveys the current emotion to the generative AI means, which then generates an appropriate message. For example, if the user's emotion is "happy," the message generated is, "Here are products related to the smartphone you are looking for!"

[0659] 3. Product Search: Information related to the product the user is looking for is retrieved from the database and displayed along with the generated message, for example, related smartphone accessories or other models.

[0660] Specific examples

[0661] If a user types in that they are looking for a "smartphone," and their emotion is "happy," the message displayed will read, "Here are products related to the smartphone you are looking for!" If their emotion is "sad," the message displayed will read, "Cheer up! We also recommend this smartphone."

[0662] Prompt Sentence Examples

[0663] “Provide relevant information about the products users are looking for and tailor your message to their emotions. If the user is happy, generate a message introducing related products. If the user is sad, generate a comforting message.”

[0664] In this way, the emotion-aware shopping assistant provides appropriate product information according to the user's emotions, realizing a more personalized shopping experience.

[0665] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0666] Step 1:

[0667] The user inputs a product search into the terminal.

[0668] Input: The keyword for the product the user is looking for (e.g., "smartphone").

[0669] Specific behavior: The user types "smartphone" into the search bar on their smartphone and presses the search button.

[0670] Output: The entered keyword is sent to the server.

[0671] Step 2:

[0672] The server uses a sentiment analysis library to analyze the user's input text and identify the sentiment.

[0673] Input: User input text (e.g., "I want this smartphone").

[0674] What happens: The server uses a sentiment analysis library (e.g. textblob or nltk) to parse the input text and identify the sentiment as "happy".

[0675] Output: Identified emotion information (e.g. "happy").

[0676] Step 3:

[0677] The server uses the emotion engine means to transmit the identified emotion information to the generative AI means.

[0678] Input: Identified emotion information (e.g. "happy").

[0679] Specific operation: The server uses the emotion engine means to send emotion information to the generative AI means.

[0680] Output: Emotional information is transmitted to the generative AI means.

[0681] Step 4:

[0682] A generative AI means generates an appropriate message based on emotional information.

[0683] Input: Emotion information (e.g., "happy") and user's search keywords (e.g., "smartphone").

[0684] Specific operation: The generative AI method generates a prompt sentence based on the emotion information and search keywords, and sends it to the AI ​​model (e.g., GPT-3). The AI ​​model generates a message such as, "Here are products related to the smartphone you are looking for!"

[0685] Output: The generated message (e.g., "Here are the smartphone-related products you're looking for!").

[0686] Step 5:

[0687] The server retrieves information related to the product the user is looking for from a database.

[0688] Input: User's search keyword (e.g. "smartphone").

[0689] What happens: The server searches the database to retrieve relevant smartphone accessories and other models.

[0690] Output: Related product information (e.g. smartphone accessories and other models).

[0691] Step 6:

[0692] The server transmits the generated message and related product information to the user's terminal.

[0693] Input: A generated message (e.g., "Here are some products related to the smartphone you're looking for!") and related product information.

[0694] Specific operation: The server sends the generated message and related product information to the user's terminal.

[0695] Output: Message and product information displayed on the user's device.

[0696] Step 7:

[0697] The user checks the message and product information displayed on the device.

[0698] Input: Message and product information sent from the server.

[0699] Specific behavior: The user checks the message and product information displayed on the device and selects a product if necessary.

[0700] Output: User selected product information.

[0701] 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.

[0702] 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.

[0703] Another example of generative AI is Gemini (registered trademark) (Internet search engine). <url: https: gemini.google.com ?hl="ja">) are mentioned.

[0704] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0705] [Second embodiment]

[0706] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0707] 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.

[0708] 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).

[0709] 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.

[0710] 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.

[0711] 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).

[0712] 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.

[0713] 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.

[0714] 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.

[0715] 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.

[0716] 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.

[0717] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0718] "Example 1"

[0719] In one embodiment of the present invention, generative AI uses natural language processing technology to understand a user's interests and preferences and generate messages based on them. The generated messages range from introducing new content the user is not aware of to responding to emotions and desires the user cannot verbalize. The generative AI learns from user responses and adjusts the message generation accordingly.

[0720] "Example 2"

[0721] As a specific example, if a generative AI determines that a user is interested in music, it will generate messages introducing new artists and songs that the user has not yet heard of. If the user responds positively to the message, the generative AI will learn from that reaction and generate similar messages more frequently. On the other hand, if the user responds negatively, the generative AI will learn from that reaction and generate similar messages less frequently.

[0722] "Example 3"

[0723] Furthermore, if generative AI determines that a user is looking for a specific product, it generates messages that introduce information related to that product or other products in the same category, thus saving users the trouble of searching for the product themselves and making it easier for them to access unknown content.

[0724] The processing flow of each embodiment will be described below.

[0725] "Example 1"

[0726] Step 1: Generative AI uses natural language processing techniques to understand the user's interests and preferences.

[0727] Step 2: The generative AI generates a message based on the user's understood interests and preferences.

[0728] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[0729] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[0730] "Example 2"

[0731] Step 1: The generative AI determines that the user is interested in music.

[0732] Step 2: Based on that judgment, the generative AI generates a message introducing new artists and songs that the user is not yet familiar with.

[0733] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[0734] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[0735] "Example 3"

[0736] Step 1: The generative AI determines that the user is looking for a specific product.

[0737] Step 2: Based on that judgment, the generative AI generates a message introducing information related to the product and other products in the same category.

[0738] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[0739] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[0740] Example 1

[0741] 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."

[0742] Conventional generative AI systems have had difficulty fully understanding a user's interests and preferences and generating appropriate messages based on them. Furthermore, they lacked the ability to learn user responses and adjust message generation accordingly, limiting the improvement of user experience. Furthermore, it was difficult to generate messages that responded to emotions and desires that users could not verbalize. To solve these issues, a system is needed that can accurately analyze a user's interests and preferences, generate appropriate messages based on those, and learn user responses to adjust message generation.

[0743] 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.

[0744] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, a means for inputting information about the user's interests and preferences, a means for analyzing the input information, a means for generating prompts based on the analysis results, a means for generating messages based on the generated prompts, and a means for collecting user responses and feeding the collected responses back to the generative AI means. This makes it possible to accurately understand the user's interests and preferences, generate appropriate messages based on the understanding, and adjust message generation accordingly by learning from the user's responses.

[0745] "Generative AI means" refers to artificial intelligence technology for generating and sending messages to users.

[0746] "Means for guiding users to services" refers to a function for guiding users to specific services or actions based on messages sent by generative AI means.

[0747] "Means for inputting information about the user's interests and preferences" refers to an interface or function that allows the user to input information about their interests and preferences into the system.

[0748] "Means for analyzing input information" refers to a function for analyzing information input by a user using natural language processing technology, etc., and identifying the user's interests and preferences.

[0749] "Means for generating prompt sentences" refers to the function of creating sentences to instruct the generative AI means to generate messages based on the analysis results.

[0750] The "means for generating a message" refers to a function for generating an appropriate message for the user based on the generated prompt sentence.

[0751] "Means for collecting user responses and providing feedback to the generative AI means" refers to a function for collecting responses given by users to generated messages and providing that information to the generative AI means for learning.

[0752] This invention is a system that generates and provides messages to users based on their interests and preferences. The system is implemented using a server, a terminal, a generative AI model, and a natural language processing library (e.g., spaCy, NLTK).

[0753] First, a user accesses the system using a terminal and inputs information about their interests and preferences. For example, the user inputs keywords such as "travel" or "cooking." This information is then sent from the terminal to the server.

[0754] The server then analyzes the received user information using a natural language processing library. Based on the analysis results, the server identifies the user's interests and preferences. For example, if the keyword "travel" is included, the server determines that the user is interested in travel.

[0755] The server then uses the analysis results to create a prompt for the generative AI model to generate a message. For example, it generates a prompt such as, "The user seems to have recently become interested in traveling. Please generate a message that includes suggestions for new travel destinations."

[0756] The generative AI model generates a message that matches the user's interests and preferences based on the prompt it receives. For example, it might generate a message like, "Hello! I heard you've recently become interested in traveling. How about a recommended travel destination, Tuscany in Italy? You can enjoy beautiful scenery and delicious food."

[0757] The generated message is sent from the server to the user's terminal, which displays the message to the user. The user can then react to the generated message, for example, by rating the message or providing feedback. This reaction is then sent from the terminal to the server.

[0758] The server collects user responses and provides them as feedback to the generative AI model. The generative AI model learns from this feedback and reflects it in future message generation. For example, if a user rates the "Tuscany" suggestion highly, the model is more likely to suggest similar travel destinations next time.

[0759] In this way, the system understands the user's interests and preferences, and generates and delivers appropriate messages based on them. By learning the user's reactions, the system can improve the accuracy of message generation in future messages.

[0760] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0761] Step 1:

[0762] Input of user interests and preferences

[0763] Users use a terminal to input information about their interests and preferences into the system, for example, by entering keywords such as "travel" or "cooking."

[0764] Input: Keywords entered by the user about their interests and preferences.

[0765] Output: The entered keyword is sent from the terminal to the server.

[0766] Step 2:

[0767] User data analysis

[0768] The server analyzes the received information about the user's interests and preferences using a natural language processing library (e.g., spaCy, NLTK).

[0769] Input: The keyword entered by the user.

[0770] Output: Data that identifies the user's interests and preferences as a result of the analysis. For example, if the keyword "travel" is included, it is determined that the user is interested in travel.

[0771] Step 3:

[0772] Generate prompt statement

[0773] Based on the analysis results, the server creates a prompt sentence for the generative AI model to generate a message.

[0774] Input: User interests and preferences identified as a result of the analysis.

[0775] Output: A prompt to send to the generative AI model. For example, "The user seems to have recently become interested in traveling. Please generate a message with suggestions for new travel destinations."

[0776] Step 4:

[0777] Message Generation

[0778] The generative AI model generates a message that matches the user's interests and preferences based on the prompt text received.

[0779] Input: The prompt text sent by the server.

[0780] Output: A generated message to be presented to the user. For example, "Hello! I heard you've been interested in traveling lately. How about a recommended destination, Tuscany, Italy? You'll enjoy beautiful scenery and delicious food."

[0781] Step 5:

[0782] Providing a message

[0783] The server sends the generated message to the user's terminal.

[0784] Input: The generated message from the generative AI model.

[0785] Output: The message sent to the user's terminal.

[0786] Step 6:

[0787] Collecting user responses

[0788] Users can react to the generated messages, for example by rating the messages and providing feedback.

[0789] Input: User response (ratings and feedback).

[0790] Output: User responses sent from the device to the server.

[0791] Step 7:

[0792] Training generative AI models

[0793] The server collects user responses and provides them as feedback to the generative AI model.

[0794] Input: User response data.

[0795] Output: The generative AI model learns and reflects this in future message generation. For example, if a user rates the "Tuscany" suggestion highly, the model will likely suggest similar travel destinations next time.

[0796] (Application example 1)

[0797] 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."

[0798] Conventional content recommendation systems have difficulty fully understanding a user's interests and preferences and recommending appropriate content. It is also difficult to generate messages that address emotions and desires that users cannot verbalize. Furthermore, they lack the ability to continuously adjust messages based on user responses, making it difficult to improve the user experience. To solve these issues, a system is needed that can deeply understand a user's interests and preferences and recommend new content based on that understanding.

[0799] 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.

[0800] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, and a means for understanding the user's interests and preferences and recommending new content based on those. This makes it possible to recommend appropriate content based on the user's interests and preferences, and to generate messages that respond to emotions and desires that the user cannot verbalize. Furthermore, by continuously adjusting the messages based on the user's reactions, the user experience can be improved.

[0801] "Generative AI methods" are artificial intelligence methods that use natural language processing technology to understand a user's interests and preferences and generate messages based on them.

[0802] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[0803] The "means for understanding the user's interests and preferences" refers to a means for analyzing the user's past behavioral data and input information to understand the user's interests and preferences.

[0804] The "means for recommending new content" is a means for selecting and recommending new content that the user does not yet know based on the user's interests and preferences.

[0805] "Means for adjusting messages based on user responses" refers to a means by which the generative AI means learns feedback from users and reflects it in future message generation.

[0806] A system for implementing this invention includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for understanding the user's interests and preferences and recommending new content based on those interests and preferences. Specific embodiments are described below.

[0807] The server uses OpenAI's API as a generative AI means. It analyzes the user's past behavioral data and input information to understand the user's interests and preferences. This analysis is performed using programming languages ​​such as Python. To recommend new content based on the user's interests and preferences, the generative AI means generates prompts and generates messages using OpenAI's API.

[0808] As a concrete example, if the user's interests are "sci-fi," "action," and "comedy," the generated recommendation message would look like this:

[0809] "Based on your interests, we recommend the new movie 'Interstellar,' the music 'Action Movie Soundtracks,' and the article 'Latest Comedy Movie Reviews.'"

[0810] Example prompt sentence:

[0811] "The user's interests and preferences are ['sci-fi', 'action', 'comedy']. Recommend new movies, music, and articles based on this."

[0812] To tailor messages based on user responses, the generative AI method learns from user feedback and incorporates it into future message generation. This feedback is learned using machine learning algorithms.

[0813] This system makes it possible to recommend appropriate content based on the user's interests and preferences, and to generate messages that respond to emotions and desires that the user cannot verbalize. Furthermore, by continuously adjusting messages based on the user's reactions, it is possible to improve the user experience.

[0814] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0815] Step 1:

[0816] The server obtains the user's interests and preferences.

[0817] Input: User ID

[0818] Output: User's interests and preferences (e.g. "sci-fi", "action", "comedy")

[0819] Specific operation: The server retrieves interest and preference data corresponding to the user ID from the database. This data is based on the user's past behavior history and input information.

[0820] Step 2:

[0821] The server generates prompts based on the user's interests and preferences.

[0822] Input: User interests and preferences

[0823] Output: Prompt (e.g., "The user's interests and preferences are ['sci-fi', 'action', 'comedy']. Based on this, please recommend new movies, music, and articles.")

[0824] Specific operation: Based on the acquired user interests and preferences, the server generates a prompt sentence to be input into the generative AI model.

[0825] Step 3:

[0826] The server generates recommendation messages using a generative AI model.

[0827] Input: prompt statement

[0828] Output: Recommendation message (e.g., "Based on your interests, we recommend the new movie 'Interstellar,' the music 'Action Movie Soundtracks,' and the article 'Latest Comedy Movie Reviews.'")

[0829] Specific operation: The server sends the generated prompt sentence to OpenAI's API and generates a recommendation message using the generative AI model.

[0830] Step 4:

[0831] The server sends the generated recommendation message to the user.

[0832] Input: Recommendation message, user ID

[0833] Output: Recommendation message displayed on the user's device

[0834] Specific operation: The server sends the generated recommendation message to the user's terminal, allowing the user to receive the message.

[0835] Step 5:

[0836] The user sends a response to the recommendation message to the server.

[0837] Input: User response (e.g. clicks, feedback comments)

[0838] Output: User response data sent to the server

[0839] Specific operation: The user clicks or makes a feedback comment on the recommendation message, and sends the reaction data to the server.

[0840] Step 6:

[0841] The server learns the user's reaction data and reflects it in subsequent message generation.

[0842] Input: User response data

[0843] Output: An updated generative AI model

[0844] Specific operation: The server analyzes the user's reaction data and adjusts the parameters of the generation AI model to reflect the results in future message generation. This enables the server to generate messages that are more suited to the user's interests and preferences.

[0845] Example 2

[0846] 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."

[0847] Conventional generative AI systems have had difficulty accurately grasping user interests and effectively recommending new content. Furthermore, they lacked a means to appropriately provide feedback on user reactions and optimize the generative AI model, making it difficult to increase user satisfaction.

[0848] 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.

[0849] In this invention, the server includes means for collecting user behavior data, means for analyzing the collected data to determine the user's interests, a generation AI means for generating messages recommending new content based on the user's interests, means for collecting user responses to the generated messages, and means for providing feedback to the generation AI means based on the collected user responses to allow it to learn. This makes it possible to accurately grasp the user's interests and recommend appropriate content. Furthermore, optimizing the generation AI model based on the user's responses can increase user satisfaction.

[0850] "User behavior data" refers to information such as the operations and selections a user makes within the system, playback history, and ratings.

[0851] "Means for collection" refers to functions and devices for acquiring and storing user behavioral data.

[0852] "Means for analyzing and determining user interests" refers to functions and devices for analyzing collected user behavior data and identifying genres and content in which the user is interested.

[0853] "Generative AI means" refers to artificial intelligence models or algorithms that generate messages recommending new content based on a user's interests.

[0854] "Means for collecting user responses" refers to a function or device for recording and storing user responses (e.g., positive and negative responses) to generated messages.

[0855] "Means of feedback and learning" refers to functions and devices that adjust the generative AI means based on collected user responses and retrain the model.

[0856] "New content" refers to information about music, artists, videos, articles, and other things that users may not yet know about or that may be of interest to them.

[0857] "Message" refers to text or notifications generated to recommend new content to users.

[0858] MODE FOR CARRYING OUT THE INVENTION

[0859] The present invention is a system that collects and analyzes user behavior data to determine the user's interests and generates messages recommending new content. A specific embodiment of this system will be described below.

[0860] Collecting user behavior data

[0861] The server uses the APIs of music streaming services and other data sources to collect user behavior data. For example, it uses APIs of Spotify and Apple Music to obtain user playback history and rating data. This allows it to understand what kind of music users listen to and which artists they like.

[0862] User interest analysis

[0863] The server analyzes the collected behavioral data to determine the user's interests. Specifically, it uses a data analysis algorithm to identify the genres and artists the user frequently listens to. Based on the analysis results, the server understands the user's musical preferences.

[0864] Generating recommendation messages for new content

[0865] The server uses a generative AI model to generate messages that recommend new content based on the user's interests. The generative AI model receives prompt sentences that reflect the user's interests as input and generates messages based on them.

[0866] As a concrete example, the following prompt sentence is input to the generative AI model:

[0867] The genre that the user has been listening to recently is "Rock," and they especially like "Queen." Please generate a message that introduces new rock artists and songs that the user has not yet heard of.

[0868] Based on this prompt, the generative AI model generates a message like this:

[0869] "Hi! I see you've been listening to a lot of Queen lately. If you haven't heard them yet, I recommend The Black Keys. I think you'll love 'Lonely Boy'!"

[0870] Collecting user responses

[0871] The device collects user reactions to the generated messages. When a user takes an action such as "Like" or "Skip" on a message, the device records the reaction data and sends it to the server.

[0872] Feedback and training for generative AI models

[0873] The server feeds back the generated AI model based on the collected user reaction data and trains it. This allows the generated AI model to optimize recommendation messages to match the user's preferences. By increasing the frequency of messages with a high percentage of positive reactions and decreasing the frequency of messages with a high percentage of negative reactions, user satisfaction can be increased.

[0874] In this way, a system is realized that collects and analyzes user behavioral data, uses a generative AI model to generate messages recommending new content, and optimizes the model based on feedback from users.

[0875] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0876] Step 1:

[0877] The server collects user behavior data. Specifically, it uses the music streaming service's API to obtain the user's playback history and rating data. It uses the user's authentication information and API key as input and obtains the user's playback history data as output. The server stores this data in a database.

[0878] Step 2:

[0879] The server analyzes the collected data to determine the user's interests. Specifically, it uses a data analysis algorithm to identify the genres and artists the user frequently listens to. As input, it uses the playback history data stored in the database, and as output, it obtains an analysis result indicating the user's interests. The server uses this analysis result in the next step.

[0880] Step 3:

[0881] The server uses the generative AI model to generate a message recommending new content based on the user's interests. Specifically, it generates a prompt sentence that reflects the user's interests and inputs it into the generative AI model. The prompt sentence created based on the analysis results is used as input, and a recommendation message is obtained as output. The server then sends this message to the user.

[0882] Step 4:

[0883] The device collects the user's reaction to the generated message. Specifically, it records whether the user performed an action such as "like" or "skip" on the message. The device uses the user's operation information as input and obtains the user's reaction data as output. The device then transmits this reaction data to the server.

[0884] Step 5:

[0885] The server feeds back the generative AI model based on the collected user reaction data and trains it. Specifically, it classifies reactions into positive and negative and provides feedback to the generative AI model. The user reaction data is used as input, and an optimized generative AI model is obtained as output. The server uses this optimized model to generate the next message.

[0886] (Application example 2)

[0887] 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."

[0888] Conventional generative AI systems have had difficulty recommending appropriate content based on a user's interests and preferences, which has led to the issue of being unable to increase user satisfaction. Additionally, they lacked a mechanism for appropriately utilizing user responses as feedback and reflecting them in subsequent recommendations, making it difficult to continuously provide content that matches the user's interests.

[0889] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for recording user responses and adjusting the frequency of message generation based on the responses. This makes it possible to recommend content based on the user's interests and preferences, thereby increasing user satisfaction. Furthermore, by appropriately utilizing the user's responses as feedback and reflecting them in the next recommendation, it becomes possible to continuously provide content that matches the user's interests.

[0890] "Generative AI means" refers to artificial intelligence technology that generates messages based on a user's interests and preferences and sends them to the user.

[0891] "Means for guiding users to services" refers to a mechanism for guiding users to specific services or content based on messages sent by generative AI means.

[0892] "Means for recording user responses" refers to technology that records as data the responses that users show to messages sent by generative AI means.

[0893] "Means for adjusting the frequency of message generation" refers to a mechanism by which the generative AI means dynamically changes the frequency at which messages are generated based on recorded user responses.

[0894] "Content" refers to the information and entertainment that users consume, such as music, videos, and articles.

[0895] "Recommendation" refers to the act of suggesting new content based on a user's interests and preferences.

[0896] A system for implementing this invention includes a generation AI means for sending messages to users, a means for directing users to services based on the messages sent by the generation AI means, and a means for recording user responses and adjusting the frequency of message generation based on the responses.

[0897] The server uses a generative AI means to generate messages based on the user's interests and preferences and sends them to the user's device. This generative AI means is implemented using a programming language such as Python and recommends appropriate content by analyzing the user's past behavioral and reaction data.

[0898] When a user receives a message and responds positively or negatively to it, the server records the response. This record is saved in a data format such as JSON. The user's response data is used as feedback when generating the next message, and the generative AI method adjusts the frequency of message generation.

[0899] As a concrete example, consider a user using a music streaming service. A generative AI tool generates a message recommending new artists or songs that the user has not yet heard of and sends it to the user's smartphone. If the user responds positively to the message, the server records the response and increases the frequency of similar recommendations next time. On the other hand, if the user responds negatively, the server decreases the frequency of similar recommendations.

[0900] Examples of prompts include:

[0901] "When a user is recommended 'Artist A - Song 1', record the action taken if they respond positively."

[0902] In this way, the server can continually provide content based on the user's interests and preferences, increasing user satisfaction.

[0903] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0904] Step 1:

[0905] The server collects users' past behavioral and reaction data, including the content they have viewed, the content they have rated, their search history, etc. The input is the user's behavioral data, and the output is a list of the collected data.

[0906] Step 2:

[0907] The server analyzes the collected data to identify the user's interests and preferences. Specifically, it uses machine learning algorithms to extract patterns of content that the user prefers. The input is a list of collected data, and the output is a profile that indicates the user's interests and preferences.

[0908] Step 3:

[0909] The server uses generative AI tools to generate messages based on the user's profile, including recommendations for new content the user does not yet know. The input is the user's profile, and the output is the generated message.

[0910] Step 4:

[0911] The server sends the generated message to the user's terminal. The input is the generated message, and the output is the message displayed on the user's terminal.

[0912] Step 5:

[0913] The user responds positively or negatively to the message displayed on the terminal. The input is the user's reaction, and the output is the reaction data.

[0914] Step 6:

[0915] The server records the user's reaction data. The input is the user's reaction data, and the output is a list of the recorded reaction data.

[0916] Step 7:

[0917] The server provides feedback based on the recorded reaction data when generating the next message. Specifically, if there are many positive reactions, it increases the number of similar recommendations, and if there are many negative reactions, it decreases the number of recommendations. The input is a list of recorded reaction data, and the output is the adjusted message generation frequency.

[0918] Example 3

[0919] 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."

[0920] In conventional systems, when users wanted to find a specific product, they had to search for it themselves, which was time-consuming and labor-intensive. It was also difficult to provide appropriate information for content that users could not verbalize or were completely unfamiliar with. Furthermore, the system lacked the ability to adjust messages based on user responses, leaving a need for an improved user experience.

[0921] 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.

[0922] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, a means for collecting user behavioral data, a means for analyzing the collected behavioral data to determine the user's intention, a means for generating a prompt sentence based on the user's intention, a means for inputting the generated prompt sentence into a generative AI model to generate a message, and a means for displaying the generated message on the user's device. This eliminates the need for users to search for information themselves and provides appropriate information even for content that cannot be verbalized or is completely unknown. Furthermore, adjusting the message based on the user's response can improve the user experience.

[0923] "Generative AI means" refers to means that use artificial intelligence technology to generate and send messages to users.

[0924] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[0925] The "means for collecting behavioral data" refers to a means for collecting behavioral data such as a user's search history and browsing history.

[0926] "Means for analyzing behavioral data to determine user intent" refers to means for analyzing collected behavioral data to determine what the user is looking for and what their intent is.

[0927] The "means for generating a prompt sentence" is a means for generating a prompt sentence to be input to a generative AI model based on the user's intention.

[0928] "Means for inputting into a generative AI model to generate a message" refers to means for inputting the generated prompt sentence into a generative AI model to generate a message introducing related information or other products in the same category.

[0929] The "means for displaying a message on a user's terminal" refers to a means for transmitting the generated message to the user's terminal and displaying it.

[0930] This invention relates to a system that, when it determines that a user is looking for a specific product, generates a message introducing information related to that product or other products in the same category. The system includes generative AI means, means for collecting user behavioral data, means for analyzing the behavioral data to determine the user's intention, means for generating a prompt sentence, means for inputting the prompt sentence into a generative AI model to generate a message, and means for displaying the generated message on the user's device.

[0931] The server collects the user's search history and browsing history using a database management system (e.g., MySQL). The collected data is analyzed using machine learning algorithms (e.g., random forests and support vector machines) to determine whether the user is looking for a specific product. For example, if it determines that the user is looking for a "smartphone," the server generates a prompt like the following:

[0932] "User is searching for a smartphone. Generate a message that provides relevant information and introduces other smartphones in the same category."

[0933] The generated prompt is then fed into a generative AI model (e.g., OpenAI's GPT-4) to generate a message introducing related information or other products in the same category. The generated message is then sent to the user's device and displayed on a web page or application interface.

[0934] For example, if the user is searching for a "smartphone," the server generates a message like this:

[0935] "Here's information related to the phone you're looking for. We also recommend the following phones: iPhone 13, Samsung Galaxy S21, Google® Pixel 6."

[0936] This system saves users the trouble of searching for information themselves and provides appropriate information for content that they cannot verbalize or are completely unfamiliar with. Furthermore, by adjusting messages based on the user's reactions, the user experience can be improved. The flow of the identification process in Example 3 will be described with reference to FIG. 15.

[0937] Step 1:

[0938] The server collects user behavior data. Specifically, it obtains information about the search keywords and products viewed by users on websites and applications from a database management system (e.g., MySQL). The input is the user's search history and browsing history, and the output is the collected behavioral data.

[0939] Step 2:

[0940] The server analyzes the collected behavioral data to determine the user's intent. Specifically, it uses machine learning algorithms (e.g., random forests and support vector machines) to analyze the user's behavioral patterns. The input is the collected behavioral data, and the output is a determination of whether the user is looking for a specific product.

[0941] Step 3:

[0942] The server generates a prompt based on the user's intent. For example, if it determines that the user is looking for a "smartphone," it generates the following prompt: "The user is looking for a smartphone. Please generate a message introducing related information and other smartphones in the same category." The input is the result of the user's intent determination, and the output is the generated prompt.

[0943] Step 4:

[0944] The server inputs the generated prompt sentence into a generative AI model to generate a message. Specifically, the prompt sentence is input into a generative AI model (e.g., OpenAI's GPT-4) to generate a message introducing related information or other products in the same category. The input is the generated prompt sentence, and the output is the generated message.

[0945] Step 5:

[0946] The server displays the generated message on the user's terminal. Specifically, it sends the generated message to the user's terminal and displays it on a web page or application interface. The input is the generated message, and the output is the message displayed on the user's terminal.

[0947] Step 6:

[0948] The user checks the displayed message and clicks to check detailed information as necessary. Specifically, when the user clicks on the displayed message, detailed information is displayed. The input is the user's click operation, and the output is the display of detailed information.

[0949] (Application example 3)

[0950] 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."

[0951] On conventional online shopping sites, when users search for a specific product, they have to search for related products or other products in the same category, which is time-consuming. It is also difficult for users to find the right product for needs they cannot verbalize or content they are completely unfamiliar with. Furthermore, there is a lack of functionality to adjust messages based on user responses, leaving a need for an improved user experience.

[0952] 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.

[0953] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for analyzing the user's search query and recommending related products or other products in the same category. This eliminates the need for the user to perform a search themselves and automatically recommends related products or other products in the same category. Furthermore, it is possible to recommend appropriate products for needs that the user cannot verbalize or content that the user is completely unfamiliar with, and by adjusting the messages based on the user's responses, the user experience can be improved.

[0954] "Generative AI methods" are artificial intelligence technologies that automatically generate related information and products based on user behavior and search queries and send them as messages.

[0955] "Means for guiding users to services" is a function for guiding users to specific services or product pages based on messages sent by generative AI means.

[0956] "Means for analyzing search queries" refers to technology that analyzes search keywords and phrases entered by users to understand their intent and relevance.

[0957] "Means for recommending related products" is a function that automatically selects and recommends products that may interest a user based on the user's search queries and behavioral history.

[0958] The "means for recommending other products in the same category" is a function for automatically selecting and recommending other products that belong to the same category as the product searched for by the user.

[0959] The system for implementing the present invention has a function of analyzing a user's search query and recommending related products or other products in the same category. A specific embodiment of this system is described below.

[0960] System Configuration

[0961] The system consists of the following main components:

[0962] 1. Generative AI methods: Artificial intelligence technology that automatically generates relevant information and products based on user behavior and search queries and sends them as messages.

[0963] 2. Means of guiding users to services: A function to guide users to specific services or product pages based on messages sent by generative AI means.

[0964] 3. Search query analysis: Techniques for analyzing search keywords and phrases entered by users to understand their intent and relevance.

[0965] 4. A means of recommending related products: A feature that automatically selects and recommends products that may be of interest to users based on their search queries and behavioral history.

[0966] 5. A means of recommending other products in the same category: A function that automatically selects and recommends other products that belong to the same category as the product searched by the user.

[0967] Hardware and software used

[0968] Hardware: Smartphones, servers

[0969] Software: Python, OpenAI API

[0970] Data processing and calculation

[0971] The server receives the search query entered by the user and inputs it as a prompt into the generative AI model. The generative AI model generates a message recommending related products or other products in the same category based on the prompt. The generated message is sent to the user's smartphone, where the user can view it.

[0972] Specific examples

[0973] If a user searches for "running shoes" on their smartphone, the server receives this search query and inputs the following prompt sentence into the generative AI model:

[0974] Example prompt sentence:

[0975] User is searching for 'running shoes'. Recommend related products or other products in the same category.

[0976] Based on this prompt, the generative AI model generates a recommendation message like the following:

[0977] An example of the message generated:

[0978] Looking for running shoes? We also recommend:

[0979] Running wear

[0980] Fitness tracker

[0981] Running socks

[0982] Sports Watch

[0983] This message is sent to the user's smartphone, allowing them to easily find related products or other products in the same category, saving the user the trouble of searching for them themselves and improving convenience.

[0984] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[0985] Step 1:

[0986] The user searches for a specific product on their device (smartphone).

[0987] Input: The search query entered by the user (e.g., "running shoes")

[0988] Output: The search query is sent to the server.

[0989] Specific action: A user types "running shoes" into the search bar on their smartphone and presses the search button.

[0990] Step 2:

[0991] The server receives and parses the search query.

[0992] Input: The search query submitted by the user

[0993] Output: Search query analysis results (user intent and relevance)

[0994] Specific operation: The server receives the search query "running shoes" and analyzes it using natural language processing techniques.

[0995] Step 3:

[0996] The server generates a prompt based on the analysis results.

[0997] Input: Analysis result of search query

[0998] Output: Generated prompt statement

[0999] Specific behavior: The server generates a prompt sentence such as "The user is looking for 'running shoes'. Please recommend related products or other products in the same category."

[1000] Step 4:

[1001] The server inputs a prompt sentence into the generative AI model, which generates a message recommending related products or other products in the same category.

[1002] Input: Generated prompt statement

[1003] Output: Recommendation message

[1004] Specific operation: The server uses the OpenAI API to input a prompt sentence into the generative AI model and generate a recommendation message.

[1005] Step 5:

[1006] The server transmits the generated recommendation message to the user's terminal.

[1007] Input: Recommendation message

[1008] Output: A recommendation message that is displayed on the user's device.

[1009] Specific operation: The server sends the generated recommendation message to the user's smartphone, allowing the user to view the message.

[1010] Step 6:

[1011] The user views the recommendation messages and selects the product of interest.

[1012] Input: Recommendation message

[1013] Output: The product page selected by the user

[1014] Specific behavior: The user checks the recommendation message on their smartphone and taps on the product they are interested in to go to the detail page.

[1015] Step 7:

[1016] The server records the user's responses and accumulates the data to improve the accuracy of recommendations in future.

[1017] Input: User selected data

[1018] Output: Accumulated user behavior data

[1019] Specific operation: The server saves the user's selection data in a database and uses it to improve the accuracy of recommendations in the future.

[1020] 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.

[1021] "Example 1"

[1022] In one embodiment of the present invention, a system includes a generative AI means including an emotion engine that recognizes a user's emotion. The emotion engine infers the user's emotion from the user's facial expression, tone of voice, message context, etc. For example, if the user expresses joy, the emotion engine conveys that information to the generative AI means, which then generates a joy-related message based on that information.

[1023] "Example 2"

[1024] Another embodiment of the present invention is a system in which an emotion engine generates a message based on a user's emotion. Specifically, when a user feels sad, the emotion engine transmits that information to a generative AI means, which then generates a comforting message based on that information.

[1025] "Example 3"

[1026] Furthermore, one embodiment of the present invention is a system in which an emotion engine adjusts messages in response to changes in the user's emotions. For example, if the user initially expresses joy but suddenly expresses sadness, the emotion engine conveys this information to the generative AI means, which then adjusts the message based on this information and generates a comforting message.

[1027] The processing flow of each embodiment will be described below.

[1028] "Example 1"

[1029] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[1030] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[1031] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotions.

[1032] "Example 2"

[1033] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[1034] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[1035] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotions. In particular, if the user feels sad, the generative AI means generates a comforting message.

[1036] "Example 3"

[1037] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[1038] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[1039] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotion. In particular, if the user's emotion changes, the generative AI means adjusts the message accordingly.

[1040] Example 1

[1041] 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."

[1042] Conventional generative AI systems have had difficulty accurately understanding users' interests and preferences and generating appropriate messages based on them. Furthermore, they lacked the ability to capture users' reactions in real time and adjust messages based on their emotions, making it difficult to improve the user experience.

[1043] The identification process by the identification processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes generative artificial intelligence means, means for guiding a user to a service based on a message transmitted by the generative artificial intelligence means, means for inputting information on the user's interests and preferences, means for analyzing the input information, means for capturing the user's reactions, means for analyzing the captured reactions and recognizing the user's emotions, and means for adjusting the message based on the recognized emotions. This makes it possible to accurately understand the user's interests and preferences, capture the user's reactions in real time to recognize their emotions, and adjust the message based on the emotions.

[1044] "Generative AI means" refers to AI technology that generates messages based on a user's interests and preferences.

[1045] "Means for guiding users to a service" refers to a function for guiding users to a specific service or action based on a message sent by the generative artificial intelligence means.

[1046] "Means for inputting information about user interests and preferences" refers to an interface or method by which a user provides information about their interests and preferences to the system.

[1047] "Means for analyzing input information" refers to the technology or algorithms used to analyze the information entered by the user and understand its content.

[1048] "Means for capturing user responses" refers to technologies or devices that record a user's response (e.g., facial expression, tone of voice, text) to a generated message.

[1049] "Means for recognizing a user's emotions by analyzing captured reactions" refers to technology or algorithms for analyzing a user's captured reactions and inferring the user's emotions from those reactions.

[1050] "Means for adjusting a message based on recognized emotions" refers to a function for appropriately changing or adjusting a message generated by a generative artificial intelligence means based on the recognized emotions of a user.

[1051] MODE FOR CARRYING OUT THE INVENTION

[1052] The present invention is a system that uses generative artificial intelligence to understand a user's interests and preferences and generate messages based on them. The system includes the ability to capture the user's reactions, recognize their emotions, and adjust the messages accordingly.

[1053] Hardware and software used

[1054] Hardware: Servers, devices (PCs, smartphones, etc.)

[1055] Software: Generative AI models (e.g., GPT-4), emotion engines, natural language processing libraries (e.g., spaCy, NLTK)

[1056] System Overview

[1057] 1. Input of user interests and preferences

[1058] User: A user uses a terminal to access the system and enter information about their interests and preferences, for example, "Where would you like to go on your next vacation?"

[1059] 2. Analysis of input data

[1060] Terminal: Sends the information entered by the user to the server.

[1061] Server: The server analyzes the received data using natural language processing libraries. Specifically, it extracts keywords and context from the input text and understands the user's interests and preferences.

[1062] 3. Message Creation

[1063] Server: Based on the analysis results, the server sends a prompt to the generative AI model. An example of a prompt is, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[1064] Generative AI: Generative AI receives a prompt and generates an appropriate message, such as, "Hawaii is a recommended destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[1065] 4. Capturing User Reactions

[1066] Device: The user receives the generated message. The device captures the user's reactions (facial expressions, tone of voice, message context, etc.). For example, if the user shows a happy expression after reading the message, the device captures that expression with a camera.

[1067] 5. Emotional Recognition

[1068] Terminal: Sends captured data to the server.

[1069] Server: The server uses an emotion engine to analyze the captured data and infer the user's emotions, for example, detecting the emotion of joy from the user's facial expression.

[1070] 6. Tailor your message

[1071] Server: Based on the information from the emotion engine, sends a new prompt to the generative AI. An example of a prompt is, "The user is expressing joy. Please provide additional information related to that emotion."

[1072] Generative AI: Generative AI receives new prompts and generates additional messages, such as "I'll also introduce some recommended tourist spots in Hawaii!"

[1073] Specific examples

[1074] User: A user interested in traveling comes into the system and types, "Where would you like to go on your next vacation?"

[1075] Terminal: Sends user input to the server.

[1076] Server: Using a natural language processing library, extract the keyword "travel destination" and understand the user's interest as "travel."

[1077] Server: Sends the generative AI a prompt saying, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[1078] Generative AI: Generates a message like, "Hawaii is a great destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[1079] Device: The user receives this message and shows a look of delight.

[1080] Device: The camera captures the user's facial expressions and sends the data to the server.

[1081] Server: Uses an emotion engine to detect the emotion of joy from the user's facial expression.

[1082] Server: Sends the prompt to the generative AI: "The user is expressing joy. Please provide additional information related to that emotion."

[1083] Generative AI: Generates an additional message such as, "We'll also introduce some recommended tourist spots in Hawaii!"

[1084] The above is an embodiment of the present invention.

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

[1086] Step 1:

[1087] User: A user uses a terminal to access the system and enter information about their interests and preferences, for example, "Where would you like to go on your next vacation?"

[1088] Input: Text information about the user's interests and preferences

[1089] Output: User input data

[1090] Step 2:

[1091] Terminal: Sends the information entered by the user to the server.

[1092] Input: User-entered data

[1093] Output: User input data sent to the server

[1094] Step 3:

[1095] Server: The server analyzes the received data using a natural language processing library (e.g., spaCy, NLTK). Specifically, it extracts keywords and context from the input text and understands the user's interests and preferences.

[1096] Input: User-entered data

[1097] Data processing: Text analysis using natural language processing libraries

[1098] Output: Analysis of user interests and preferences

[1099] Step 4:

[1100] Server: Based on the analysis results, the server sends a prompt to a generative AI model (e.g., GPT-4). An example of a prompt is, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[1101] Input: Analysis results about user interests and preferences

[1102] Data processing: Prompt generation

[1103] Output: Prompt for the generative AI model

[1104] Step 5:

[1105] Generative AI: Generative AI receives a prompt and generates an appropriate message, such as, "Hawaii is a recommended destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[1106] Input: prompt statement

[1107] Data processing: Message generation

[1108] Output: The generated message

[1109] Step 6:

[1110] Device: The user receives the generated message. The device captures the user's reactions (facial expressions, tone of voice, message context, etc.). For example, if the user shows a happy expression after reading the message, the device captures that expression with a camera.

[1111] Input: The generated message

[1112] Output: User response data

[1113] Step 7:

[1114] Terminal: Sends captured data to the server.

[1115] Input: User response data

[1116] Output: User response data sent to the server

[1117] Step 8:

[1118] Server: The server uses an emotion engine to analyze the captured data and infer the user's emotions, for example, detecting the emotion of joy from the user's facial expression.

[1119] Input: User response data

[1120] Data processing: Emotion analysis using an emotion engine

[1121] Output: Analysis results on user sentiment

[1122] Step 9:

[1123] Server: Based on the information from the emotion engine, sends a new prompt to the generative AI. An example of a prompt is, "The user is expressing joy. Please provide additional information related to that emotion."

[1124] Input: Analysis results on user sentiment

[1125] Data processing: generating new prompts

[1126] Output: A new prompt for the generative AI model

[1127] Step 10:

[1128] Generative AI: Generative AI receives new prompts and generates additional messages, such as "I'll also introduce some recommended tourist spots in Hawaii!"

[1129] Input: new prompt text

[1130] Data processing: generating additional messages

[1131] Output: Additional messages generated

[1132] (Application example 1)

[1133] 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."

[1134] Conventional content delivery services recommend content based on users' interests and preferences, but they lack personalized recommendations that take into account the user's current emotions. Furthermore, it is difficult to generate messages that respond to emotions and desires that users cannot verbalize. This limits the improvement of the user experience.

[1135] 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: a generative AI means for sending a message to a user; a means for guiding the user to a service based on the message sent by the generative AI means; an emotion recognition means for recognizing emotions by analyzing the user's facial expressions and tone of voice; a preference understanding means for understanding the user's interests and preferences by analyzing the user's past viewing history and search history; and a content recommendation means for recommending personalized content based on the emotion recognition means and preference understanding means. This enables personalized content recommendation based on the user's current emotions and past preferences, improving the user experience.

[1136] "Generative AI methods" are artificial intelligence methods that understand a user's interests and preferences and generate messages based on them.

[1137] "Means for guiding users to services" refers to means for guiding users to specific services or actions based on messages sent by generative AI means.

[1138] The "emotion recognition means" is a means for recognizing the user's emotions by analyzing the user's facial expressions and tone of voice.

[1139] The "preference understanding means" is a means for understanding the user's interests and preferences by analyzing the user's past viewing history and search history.

[1140] The "content recommendation means" is a means for recommending personalized content to a user based on the emotion recognition means and the preference understanding means.

[1141] A system for implementing the present invention includes the following means for recommending personalized content based on a user's emotions and preferences.

[1142] System Configuration

[1143] 1. Generative AI means:

[1144] The server uses generative AI to understand the user's interests and preferences, and generates messages based on them. The generated messages address feelings and desires that the user cannot verbalize, or introduce new content that the user is completely unfamiliar with.

[1145] 2. Means of directing you to the service:

[1146] The server guides the user to specific services or actions based on the messages sent by the generative AI means, allowing the user to access the right content or services at the right time.

[1147] 3. Emotion recognition means:

[1148] The device uses a camera and microphone to recognize emotions by analyzing the user's facial expressions and tone of voice. The emotion recognition engine analyzes this data in real time to estimate the user's current emotion.

[1149] 4. Means of understanding preferences:

[1150] The server analyzes the user's past viewing and search history to understand their interests and preferences, and uses natural language processing technology to extract these from the text entered by the user.

[1151] 5. Content recommendation methods:

[1152] The server recommends personalized content to users based on emotion recognition and preference understanding. The recommendation algorithm learns user responses and is continually adjusted.

[1153] Hardware and software used

[1154] Hardware:

[1155] Devices such as smartphones and tablets

[1156] Camera and microphone

[1157] software:

[1158] Emotion recognition engines (e.g., the sentiment-analysis pipeline in the transformers library)

[1159] Natural language processing technology (e.g., TfidfVectorizer)

[1160] Similarity calculation algorithm (e.g., cosine_similarity)

[1161] Specific examples

[1162] When a user smiles into the smartphone camera, the emotion recognition engine recognizes "joy," and the server recommends action-packed thrillers or heartwarming romantic comedies based on the action movies and comedy shows the user has watched in the past.

[1163] Prompt Sentence Examples

[1164] "Analyze the user's facial expressions and tone of voice to recognize their current emotions. Understand their interests and preferences based on their past viewing and search history, and recommend the most appropriate content."

[1165] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1166] Step 1:

[1167] The device captures the user's facial expressions and tone of voice using a camera and microphone. As input, it receives image data from the camera and audio data from the microphone. These data are sent to an emotion recognition engine to estimate the user's current emotion. As output, it obtains the user's emotion (e.g., joy, sadness, surprise, etc.).

[1168] Step 2:

[1169] The server retrieves the user's past viewing and search history from a database. It uses the user ID as input to search for related history data. It then uses natural language processing technology (e.g., TfidfVectorizer) to vectorize this history data and extract the user's interests and preferences. As an output, it obtains vector data representing the user's interests and preferences.

[1170] Step 3:

[1171] The server integrates the data obtained from the emotion recognition means and the preference understanding means. As input, it receives the user's emotion data and preference vector data. Based on this data, the generative AI means generates the optimal message for the user. As output, a personalized message is obtained.

[1172] Step 4:

[1173] The server sends the generated message to the user. As input, it receives message data from the generative AI means. The message is sent to the user's device in the form of a push notification, email, etc. As output, the user receives the message.

[1174] Step 5:

[1175] Users access services based on the messages they receive. As input, they review the message content and execute the provided links or actions, which direct the user to specific content or services. As output, they obtain data on the user's actions.

[1176] Step 6:

[1177] The server collects user reaction data and feeds it back to the generative AI means. As input, it receives user action data. Based on this data, the generative AI means adjusts the message generation algorithm. As output, it obtains an improved message generation model.

[1178] Example 2

[1179] 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."

[1180] Conventional message generation systems have difficulty generating personalized messages based on users' interests and emotions, making it difficult to increase user satisfaction. Furthermore, they lacked a mechanism for learning user responses and improving the accuracy of message generation, making it difficult to provide services that meet user needs. Furthermore, they were unable to generate appropriate messages based on users' emotions, making it impossible to respond in a way that was sensitive to the user's feelings.

[1181] 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.

[1182] In this invention, the server includes a generative artificial intelligence means for sending messages to users, a means for directing users to services based on the messages sent by the generative artificial intelligence means, a means for analyzing the user's past behavioral data to determine their interests, a means for collecting user responses and adjusting the frequency of message generation based on the responses, and a means for analyzing the user's emotions and generating messages based on the emotions. This enables the generation of personalized messages based on the user's interests and emotions, thereby increasing user satisfaction. Furthermore, the server can learn user responses to improve the accuracy of message generation, enabling the provision of services that meet the user's needs. Furthermore, by generating appropriate messages based on the user's emotions, it is possible to respond in accordance with the user's emotions.

[1183] "Generative AI means" refers to means that uses artificial intelligence technology to generate and send messages to users.

[1184] The "means for guiding a user to a service" is a means for guiding a user to a specific service or action based on a message sent by the generative artificial intelligence means.

[1185] The "means for determining interests by analyzing past behavioral data" is a means for analyzing the user's past behavioral data (for example, music playback history and search history) and identifying the user's interests.

[1186] The "means for collecting user responses and adjusting the frequency of message generation based on the responses" refers to a means for collecting user responses to generated messages and adjusting the frequency of future message generation based on the responses.

[1187] The "means for analyzing the user's emotions and generating a message based on those emotions" refers to a means for analyzing the user's emotions and generating an appropriate message in accordance with those emotions.

[1188] This invention is a system that generates personalized messages based on a user's interests and emotions, and improves the accuracy of message generation by learning the user's reactions. This system includes generative artificial intelligence means, means for directing users to services, means for analyzing past behavioral data to determine interests, means for collecting user reactions and adjusting the frequency of message generation based on the reactions, and means for analyzing the user's emotions and generating messages based on the emotions.

[1189] The server collects the user's past behavioral data and analyzes it using machine learning libraries such as TensorFlow and PyTorch to determine the genre of music the user is interested in. For example, it retrieves the user's music history from a database over the past month and identifies the music genre they listen to most frequently.

[1190] The server then uses a generative AI model (e.g., GPT-3) to generate messages introducing new artists and songs that the user might be interested in. The generated messages are customized based on the user's interests. For example, if it determines that the user is interested in rock music, it sends the generative AI model a prompt saying, "Generate a message introducing a new rock artist." The generative AI model generates the message, "Check out the new rock band, 'The New Rockers'!"

[1191] The device displays the generated message to the user. If the user responds positively or negatively to the message, the device collects the response and sends it to the server. For example, if the device displays the generated message on the user's screen and the user clicks the "Like!" button, the device sends that information to the server. Conversely, if the user clicks the "Not Interested" button, the device also sends the information.

[1192] The server receives user responses and stores them as learning data. It increases the frequency of messages with a large number of positive responses and decreases the frequency of messages with a large number of negative responses. This learning process uses a reinforcement learning algorithm. For example, the server stores messages to which users have responded with a "Like!" in a learning database, and refers to this data when generating similar messages from the next time onwards to adjust the frequency of their generation.

[1193] Furthermore, the emotion engine analyzes the user's emotions. If the emotion engine determines that the user is sad, it passes that information on to the generative AI model. The generative AI model uses that information to generate a message to comfort the user. For example, the emotion engine analyzes the user's facial expressions and input text and determines that the user is sad. The emotion engine sends that information to the server, and the server prompts the generative AI model to "generate a message to comfort the user." The generative AI model generates a message saying, "You've had a tough day today. Why don't you listen to some relaxing music?"

[1194] Example prompt sentence:

[1195] Analyze the user's recent music history and generate messages introducing new artists or songs the user hasn't heard of. If the user responds positively, learn from that reaction and generate similar messages more frequently. If the user responds negatively, learn from that reaction and generate similar messages less frequently.

[1196] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1197] Step 1:

[1198] The server collects data on the user's past behavior. Specifically, it retrieves data from a database, such as the music history the user has played, search history, and links they have clicked. This data is used as input for analysis to identify the user's interests. Machine learning libraries such as TensorFlow and PyTorch are used for the analysis to identify the music genres the user listens to most frequently. The output is data indicating the user's interests.

[1199] Step 2:

[1200] The server uses a generative AI model (e.g., GPT-3) to generate messages introducing new artists and songs that the user might be interested in. As input, it uses the user interest data obtained in step 1. Specifically, it sends the generative AI model a prompt: "Generate a message introducing a new rock music artist." The output is the generated message. For example, the message generated is "Check out the new rock band, 'The New Rockers'!"

[1201] Step 3:

[1202] The device displays the generated message to the user. It uses the message generated in step 2 as input. Specifically, the device displays the message on the screen, and the user responds positively or negatively to the message. The output is the user's response data. For example, if the user clicks the "Like" button, that information is collected by the device.

[1203] Step 4:

[1204] The device sends the user's reaction data to the server. The user's reaction data collected in step 3 is used as input. Specifically, information on whether the user clicked the "Like" button or the "Not Interested" button is sent to the server. The output is the user's reaction data sent to the server.

[1205] Step 5:

[1206] The server receives the user's reaction data and stores it as training data. It uses the user's reaction data sent in step 4 as input. Specifically, it uses a reinforcement learning algorithm to increase the frequency of messages with many positive reactions and decrease the frequency of messages with many negative reactions. The output is updated message generation frequency data.

[1207] Step 6:

[1208] The emotion engine analyzes the user's emotions. It uses the user's facial expressions and input text as input. Specifically, it uses emotion recognition software such as OpenCV or Emotion API to determine whether the user is sad. The output is the user's emotion data.

[1209] Step 7:

[1210] The server receives the user's emotion data obtained from the emotion engine and transmits it to the generative AI model. The user's emotion data obtained in step 6 is used as input. Specifically, it sends a prompt to the generative AI model saying, "Please generate a message to comfort the user." The output is the generated comforting message. For example, a message might be generated saying, "You've had a tough day today. Why don't you listen to some relaxing music?"

[1211] (Application example 2)

[1212] 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."

[1213] Conventional content delivery services have difficulty generating personalized messages based on users' interests and emotions, which has led to insufficient improvements in user experience. Furthermore, they lack the ability to learn users' reactions and adjust the frequency of message generation, making it difficult to recommend appropriate content to users.

[1214] 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.

[1215] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, an emotion engine means for analyzing user emotions, a means for generating messages based on emotion data acquired by the emotion engine means, and a means for learning user reactions and adjusting the frequency of message generation. This enables the generation of personalized messages based on the user's interests and emotions, improving the user experience and realizing appropriate content recommendations.

[1216] "Generative AI means" refers to artificial intelligence technology for generating and sending messages to users.

[1217] "Means for guiding users to services" refers to a function for guiding users to specific services or content based on messages sent by generative AI means.

[1218] "Emotion engine means" refers to a technology for analyzing a user's emotions and acquiring the emotion data.

[1219] The "means for generating a message based on emotion data" refers to a function for generating an appropriate message for the user based on emotion data acquired by the emotion engine means.

[1220] "Means for learning user reactions and adjusting the frequency of message generation" refers to a function for analyzing user reactions and increasing or decreasing the frequency of message generation based on the results.

[1221] A system for implementing the present invention generates personalized messages for users and recommends appropriate content based on the user's emotions. A specific embodiment of this system is described below.

[1222] System Configuration

[1223] 1. Generative AI means:

[1224] The server generates and sends messages to users using a generative AI means that generates messages recommending new artists and songs based on the user's interests and past behavioral data.

[1225] 2. Means of directing you to the service:

[1226] The server directs the user to specific services and content based on messages sent by the generative AI means, which introduces new music and artists that may interest the user and directs the user to music streaming services.

[1227] 3. Emotion engine means:

[1228] The server analyzes the user's emotions using an emotion engine means that analyzes the user's text data and voice data and classifies the user's emotional state into "positive," "neutral," or "negative."

[1229] 4. Means of generating messages based on emotion data:

[1230] The server generates an appropriate message for the user based on the emotion data acquired by the emotion engine means. For example, if the user is expressing negative emotions, the server generates a message of comfort or encouragement.

[1231] 5. Learn user reactions and adjust message generation frequency:

[1232] The server learns user responses and adjusts the frequency of message generation based on the results: if there are a lot of positive responses, it generates similar messages more frequently, and if there are a lot of negative responses, it generates them less frequently.

[1233] Hardware and software used

[1234] Hardware: Servers, user devices (smartphones, tablets, PCs, etc.)

[1235] Software: Generative AI models (e.g., GPT-3), sentiment analysis libraries (e.g., TextBlob), music recommendation APIs (e.g., Spotify API)

[1236] Specific examples of processing

[1237] 1. User sentiment analysis:

[1238] The server analyzes the user's text data using an emotion engine means and classifies the user's emotions into "positive," "neutral," and "negative." For example, the server uses the TextBlob library to perform the emotion analysis.

[1239] 2. Message Creation:

[1240] The server uses generative AI tools to generate messages based on the emotion data. For example, if the user is expressing negative emotions, the server generates a message like, "Feeling down? Maybe these tracks can cheer you up."

[1241] 3. Music Recommendations:

[1242] The server uses a music recommendation API to recommend new artists and songs that might suit the user, for example, using the Spotify API to make recommendations based on the user's listening history.

[1243] Prompt Sentence Examples

[1244] I am developing an application that recommends music based on the user's emotions. Assuming that the user's emotions can be classified as positive, neutral, or negative, please generate Python code to provide messages and music recommendations according to each emotion. I will use the TextBlob library for sentiment analysis and an external music API for music recommendation.

[1245] In this way, personalized messages and appropriate content recommendations based on the user's interests and emotions can be generated.

[1246] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1247] Step 1:

[1248] The server acquires the user's text data. The input is the text data entered by the user, and the output is the text data. Specifically, the server acquires comments and feedback entered by the user into the application.

[1249] Step 2:

[1250] The server uses an emotion engine to analyze the acquired text data and classify the user's emotion. The input is the text data acquired in step 1, and the output is an emotion classification of "positive," "neutral," or "negative." Specifically, the server uses the TextBlob library to perform emotion analysis of the text.

[1251] Step 3:

[1252] The server generates a message using a generative AI means based on the emotion data acquired by the emotion engine means. The input is the emotion classification obtained in step 2, and the output is the message to be sent to the user. Specifically, a generative AI model (e.g., GPT-3) is used to generate a message according to the emotion.

[1253] Step 4:

[1254] The server sends the generated message to the user. The input is the message generated in step 3, and the output is the message displayed on the user's device. Specifically, the message is notified to the user's smartphone or PC.

[1255] Step 5:

[1256] The server receives user responses. The input is the user's response to the message (e.g., click, comment, feedback), and the output is the response data. Specifically, it records what actions the user took on the message.

[1257] Step 6:

[1258] The server analyzes the acquired user reaction data and uses it as learning data for the generative AI means. The input is the reaction data acquired in step 5, and the output is the learning result. Specifically, if there are many positive reactions, it increases the frequency of generating similar messages, and if there are many negative reactions, it decreases the frequency.

[1259] Step 7:

[1260] The server uses a music recommendation API to recommend new artists and songs suitable for the user. The input is the user's playback history and interest data, and the output is recommended music content. Specifically, it uses the Spotify API to recommend music based on the user's preferences.

[1261] Step 8:

[1262] The server provides the recommended music content to the user. The input is the music content recommended in step 7, and the output is a music recommendation list that is displayed on the user's device. Specifically, the music recommendation list is displayed on the user's smartphone or PC.

[1263] In this way, personalized messages and appropriate content recommendations based on the user's interests and emotions can be generated.

[1264] Example 3

[1265] 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."

[1266] With conventional systems, it was very time-consuming for users to search for specific products, and it was difficult to provide appropriate messages that responded to changes in the user's emotions. Furthermore, information was insufficient for content that the user could not verbalize or was completely unfamiliar with. This resulted in a poor user experience and hindered the promotion of service usage.

[1267] 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.

[1268] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, a means for collecting user behavioral data, a means for analyzing the collected behavioral data, a means for generating related information based on the analysis results, a means for detecting changes in the user's emotions, and a means for adjusting messages based on the detected changes in emotions. This eliminates the need for users to search for specific products and enables the provision of appropriate messages in response to changes in the user's emotions. It also enables the provision of information about content that the user cannot verbalize or is completely unfamiliar with.

[1269] "Generative AI means" refers to means that use artificial intelligence technology to generate and send messages to users.

[1270] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[1271] The "means for collecting behavioral data" refers to a means for collecting behavioral data such as a user's search history and click history.

[1272] The "means for analyzing behavioral data" refers to a means for analyzing collected user behavioral data and identifying the user's intentions and interests.

[1273] The "means for generating related information" is a means for generating information related to the product the user is looking for or other products in the same category based on the analysis results of the behavioral data.

[1274] The "means for detecting changes in emotions" is a means for monitoring the user's inputs and actions in real time and detecting changes in the user's emotions.

[1275] The "means for adjusting the message" refers to a means for adjusting the message using generative AI means based on detected changes in the user's emotions and changing the content to appropriate content.

[1276] This invention is a system that saves users the trouble of searching for specific products and provides appropriate messages that respond to changes in the user's emotions.It also provides information about things that the user cannot put into words or about content that the user is completely unfamiliar with.

[1277] Hardware and software used

[1278] Hardware: Servers, user devices (PCs, smartphones, etc.)

[1279] Software: Generative AI models (e.g., GPT-4), emotion engines, database management systems (e.g., MySQL)

[1280] System configuration

[1281] 1. The server collects user behavior data by storing the user's search history and click history on the device in a database.

[1282] 2. The server uses machine learning algorithms to analyze the collected behavioral data, thereby identifying the products and interests of the user.

[1283] 3. The server uses a generative AI model (e.g., GPT-4) to generate messages that provide information related to the product the user is looking for and introduce other products in the same category.

[1284] 4. The emotion engine monitors user input and behavior in real time to detect changes in the user's emotions.

[1285] 5. The server uses the generative AI model to adjust the message based on the emotional change information received from the emotion engine and sends it to the user.

[1286] Specific examples

[1287] When a user is looking for a specific product

[1288] A user searches for "smartphone."

[1289] The server collects and analyzes the search history.

[1290] The server uses the generative AI model to generate and send messages to users containing relevant information, such as "latest smartphone reviews" or "recommended smartphone accessories."

[1291] Message adjustment according to user's emotional changes

[1292] The user types, "I'm so happy I bought a new smartphone."

[1293] The emotion engine detects the emotion.

[1294] The user types, "I'm sad my phone is broken."

[1295] The emotion engine detects the change in emotion and communicates it to the server.

[1296] The server uses the generative AI model to generate a comforting message to the user, such as "I see you're having trouble. Here's how to contact our support center," and sends it to the user.

[1297] Prompt Sentence Examples

[1298] Generate product-related information

[1299] “If a user is looking for a smartphone, generate a message with relevant information and other products in the same category.”

[1300] Adapting messages to reflect changing emotions

[1301] "If the user's emotion changes from happy to sad, generate a comforting message."

[1302] In this way, it is possible to save the user the trouble of searching for a specific product and to provide appropriate messages according to changes in the user's emotions. It is also possible to provide information about content that the user cannot verbalize or is completely unfamiliar with. The flow of the identification process in the third embodiment will be described with reference to FIG. 21.

[1303] Step 1:

[1304] Collecting user behavior data

[1305] Input: Search and click history of the user on their device

[1306] What it does: When a user searches for "smartphone," the search query is sent to the server, which stores this data in a database.

[1307] Output: Stored user behavior data

[1308] Step 2:

[1309] User behavior analysis

[1310] Input: Stored user behavior data

[1311] How it works: The server uses machine learning algorithms to analyze user behavior data. For example, it determines that a user is searching for a "smartphone" based on their past search and click history.

[1312] Output: Analysis of user search results and interests

[1313] Step 3:

[1314] Generating related information

[1315] Input: Analytics about the products and interests of users

[1316] Specific operation: The server inputs a prompt to the generative AI model (e.g., GPT-4) saying, "If the user is looking for a smartphone, please generate a message that introduces related information and other products in the same category." The generative AI model then generates a message based on this.

[1317] Output: Messages generated containing relevant information

[1318] Step 4:

[1319] Detecting changes in emotions

[1320] Input: User input or actions

[1321] How it works: The emotion engine monitors user input and behavior in real time. For example, if a user types "I'm happy because I bought a new smartphone" and then "I'm sad because my smartphone broke," the emotion engine will detect this change.

[1322] Output: Information on detected changes in user emotions

[1323] Step 5:

[1324] Message coordination and generation

[1325] Input: Detected changes in user emotion

[1326] Specific operation: The server inputs a prompt to the generative AI model: "When the user's emotion changes from joy to sadness, please generate a comforting message." The generative AI model then generates a message based on this.

[1327] Output: The adjusted message

[1328] The above is the specific flow of the program processing of this system.

[1329] (Application example 3)

[1330] 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."

[1331] On traditional online shopping sites, users often spend a lot of time and effort searching for specific products. Furthermore, the lack of appropriate product suggestions and messaging tailored to the user's emotions hinders the user experience. Furthermore, the lack of suggestions for needs that users cannot verbalize or for content they are unfamiliar with limits users' opportunities to discover new products and services.

[1332] 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.

[1333] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, an emotion engine means for analyzing the user's emotions, and a means for adjusting the messages based on the emotion information analyzed by the emotion engine means. This eliminates the need for users to search for specific products, enables appropriate product suggestions based on emotions, and improves the user experience. It can also make suggestions for needs that users cannot verbalize or for unknown content.

[1334] "Generative AI means" refers to artificial intelligence technology that generates and sends messages to users.

[1335] "Means for guiding users to services" refers to technology for guiding users to specific services or product pages based on messages sent by generative AI means.

[1336] "Emotion engine means" refers to a technology for analyzing a user's emotions and acquiring that emotion information.

[1337] The "means for adjusting a message" refers to a technology for appropriately changing the message generated by the generative AI means based on the emotional information analyzed by the emotion engine means.

[1338] A system for carrying out the present invention includes a generative AI means for analyzing a user's emotions and generating an appropriate message based on the emotions, a means for directing the user to a service based on the generated message, an emotion engine means for analyzing the user's emotions, and a means for adjusting the message based on the emotion information.

[1339] System Program

[1340] The system program performs the following processing.

[1341] Hardware

[1342] Smartphone

[1343] software

[1344] Python

[1345] Sentiment analysis libraries (e.g. textblob, nltk)

[1346] Generative AI models (e.g., GPT-3)

[1347] Data processing and calculation

[1348] 1. Sentiment analysis: The sentiment analysis library analyzes the user's input text to identify the emotion. For example, if a user types "I want this smartphone," the sentiment analysis library analyzes the text and identifies the user's emotion as "happy."

[1349] 2. Message generation: The emotion engine means conveys the current emotion to the generative AI means, which then generates an appropriate message. For example, if the user's emotion is "happy," the message generated is, "Here are products related to the smartphone you are looking for!"

[1350] 3. Product Search: Information related to the product the user is looking for is retrieved from the database and displayed along with the generated message, for example, related smartphone accessories or other models.

[1351] Specific examples

[1352] If a user types in that they are looking for a "smartphone," and their emotion is "happy," the message displayed will read, "Here are products related to the smartphone you are looking for!" If their emotion is "sad," the message displayed will read, "Cheer up! We also recommend this smartphone."

[1353] Prompt Sentence Examples

[1354] “Provide relevant information about the products users are looking for and tailor your message to their emotions. If the user is happy, generate a message introducing related products. If the user is sad, generate a comforting message.”

[1355] In this way, the emotion-aware shopping assistant provides appropriate product information according to the user's emotions, realizing a more personalized shopping experience.

[1356] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1357] Step 1:

[1358] The user inputs a product search into the terminal.

[1359] Input: The keyword for the product the user is looking for (e.g., "smartphone").

[1360] Specific behavior: The user types "smartphone" into the search bar on their smartphone and presses the search button.

[1361] Output: The entered keyword is sent to the server.

[1362] Step 2:

[1363] The server uses a sentiment analysis library to analyze the user's input text and identify the sentiment.

[1364] Input: User input text (e.g., "I want this smartphone").

[1365] What happens: The server uses a sentiment analysis library (e.g. textblob or nltk) to parse the input text and identify the sentiment as "happy".

[1366] Output: Identified emotion information (e.g. "happy").

[1367] Step 3:

[1368] The server uses the emotion engine means to transmit the identified emotion information to the generative AI means.

[1369] Input: Identified emotion information (e.g. "happy").

[1370] Specific operation: The server uses the emotion engine means to send emotion information to the generative AI means.

[1371] Output: Emotional information is transmitted to the generative AI means.

[1372] Step 4:

[1373] A generative AI means generates an appropriate message based on emotional information.

[1374] Input: Emotion information (e.g., "happy") and user's search keywords (e.g., "smartphone").

[1375] Specific operation: The generative AI method generates a prompt sentence based on the emotion information and search keywords, and sends it to the AI ​​model (e.g., GPT-3). The AI ​​model generates a message such as, "Here are products related to the smartphone you are looking for!"

[1376] Output: The generated message (e.g., "Here are the smartphone-related products you're looking for!").

[1377] Step 5:

[1378] The server retrieves information related to the product the user is looking for from a database.

[1379] Input: User's search keyword (e.g. "smartphone").

[1380] What happens: The server searches the database to retrieve relevant smartphone accessories and other models.

[1381] Output: Related product information (e.g. smartphone accessories and other models).

[1382] Step 6:

[1383] The server transmits the generated message and related product information to the user's terminal.

[1384] Input: A generated message (e.g., "Here are some products related to the smartphone you're looking for!") and related product information.

[1385] Specific operation: The server sends the generated message and related product information to the user's terminal.

[1386] Output: Message and product information displayed on the user's device.

[1387] Step 7:

[1388] The user checks the message and product information displayed on the device.

[1389] Input: Message and product information sent from the server.

[1390] Specific behavior: The user checks the message and product information displayed on the device and selects a product if necessary.

[1391] Output: User selected product information.

[1392] 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.

[1393] 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.

[1394] Another example of generative AI is Gemini (internet search engine). <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1395] 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.

[1396] [Third embodiment]

[1397] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1398] 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.

[1399] 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).

[1400] 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.

[1401] 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.

[1402] 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).

[1403] 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.

[1404] 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.

[1405] 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.

[1406] 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.

[1407] 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.

[1408] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1409] "Example 1"

[1410] In one embodiment of the present invention, generative AI uses natural language processing technology to understand a user's interests and preferences and generate messages based on them. The generated messages range from introducing new content the user is not aware of to responding to emotions and desires the user cannot verbalize. The generative AI learns from user responses and adjusts the message generation accordingly.

[1411] "Example 2"

[1412] As a specific example, if a generative AI determines that a user is interested in music, it will generate messages introducing new artists and songs that the user has not yet heard of. If the user responds positively to the message, the generative AI will learn from that reaction and generate similar messages more frequently. On the other hand, if the user responds negatively, the generative AI will learn from that reaction and generate similar messages less frequently.

[1413] "Example 3"

[1414] Furthermore, if generative AI determines that a user is looking for a specific product, it generates messages that introduce information related to that product or other products in the same category, thus saving users the trouble of searching for the product themselves and making it easier for them to access unknown content.

[1415] The processing flow of each embodiment will be described below.

[1416] "Example 1"

[1417] Step 1: Generative AI uses natural language processing techniques to understand the user's interests and preferences.

[1418] Step 2: The generative AI generates a message based on the user's understood interests and preferences.

[1419] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[1420] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[1421] "Example 2"

[1422] Step 1: The generative AI determines that the user is interested in music.

[1423] Step 2: Based on that judgment, the generative AI generates a message introducing new artists and songs that the user is not yet familiar with.

[1424] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[1425] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[1426] "Example 3"

[1427] Step 1: The generative AI determines that the user is looking for a specific product.

[1428] Step 2: Based on that judgment, the generative AI generates a message introducing information related to the product and other products in the same category.

[1429] Step 3: The generated message is sent to the user, and the generative AI learns the user's reaction.

[1430] Step 4: Based on the learned user responses, the generative AI adjusts the message generation accordingly.

[1431] Example 1

[1432] 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."

[1433] Conventional generative AI systems have had difficulty fully understanding a user's interests and preferences and generating appropriate messages based on them. Furthermore, they lacked the ability to learn user responses and adjust message generation accordingly, limiting the improvement of user experience. Furthermore, it was difficult to generate messages that responded to emotions and desires that users could not verbalize. To solve these issues, a system is needed that can accurately analyze a user's interests and preferences, generate appropriate messages based on those, and learn user responses to adjust message generation.

[1434] 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.

[1435] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, a means for inputting information about the user's interests and preferences, a means for analyzing the input information, a means for generating prompts based on the analysis results, a means for generating messages based on the generated prompts, and a means for collecting user responses and feeding the collected responses back to the generative AI means. This makes it possible to accurately understand the user's interests and preferences, generate appropriate messages based on the understanding, and adjust message generation accordingly by learning from the user's responses.

[1436] "Generative AI means" refers to artificial intelligence technology for generating and sending messages to users.

[1437] "Means for guiding users to services" refers to a function for guiding users to specific services or actions based on messages sent by generative AI means.

[1438] "Means for inputting information about the user's interests and preferences" refers to an interface or function that allows the user to input information about their interests and preferences into the system.

[1439] "Means for analyzing input information" refers to a function for analyzing information input by a user using natural language processing technology, etc., and identifying the user's interests and preferences.

[1440] "Means for generating prompt sentences" refers to the function of creating sentences to instruct the generative AI means to generate messages based on the analysis results.

[1441] The "means for generating a message" refers to a function for generating an appropriate message for the user based on the generated prompt sentence.

[1442] "Means for collecting user responses and providing feedback to the generative AI means" refers to a function for collecting responses given by users to generated messages and providing that information to the generative AI means for learning.

[1443] This invention is a system that generates and provides messages to users based on their interests and preferences. The system is implemented using a server, a terminal, a generative AI model, and a natural language processing library (e.g., spaCy, NLTK).

[1444] First, a user accesses the system using a terminal and inputs information about their interests and preferences. For example, the user inputs keywords such as "travel" or "cooking." This information is then sent from the terminal to the server.

[1445] The server then analyzes the received user information using a natural language processing library. Based on the analysis results, the server identifies the user's interests and preferences. For example, if the keyword "travel" is included, the server determines that the user is interested in travel.

[1446] The server then uses the analysis results to create a prompt for the generative AI model to generate a message. For example, it generates a prompt such as, "The user seems to have recently become interested in traveling. Please generate a message that includes suggestions for new travel destinations."

[1447] The generative AI model generates a message that matches the user's interests and preferences based on the prompt it receives. For example, it might generate a message like, "Hello! I heard you've recently become interested in traveling. How about a recommended travel destination, Tuscany in Italy? You can enjoy beautiful scenery and delicious food."

[1448] The generated message is sent from the server to the user's terminal, which displays the message to the user. The user can then react to the generated message, for example, by rating the message or providing feedback. This reaction is then sent from the terminal to the server.

[1449] The server collects user responses and provides them as feedback to the generative AI model. The generative AI model learns from this feedback and reflects it in future message generation. For example, if a user rates the "Tuscany" suggestion highly, the model is more likely to suggest similar travel destinations next time.

[1450] In this way, the system understands the user's interests and preferences, and generates and delivers appropriate messages based on them. By learning the user's reactions, the system can improve the accuracy of message generation in future messages.

[1451] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1452] Step 1:

[1453] Input of user interests and preferences

[1454] Users use a terminal to input information about their interests and preferences into the system, for example, by entering keywords such as "travel" or "cooking."

[1455] Input: Keywords entered by the user about their interests and preferences.

[1456] Output: The entered keyword is sent from the terminal to the server.

[1457] Step 2:

[1458] User data analysis

[1459] The server analyzes the received information about the user's interests and preferences using a natural language processing library (e.g., spaCy, NLTK).

[1460] Input: The keyword entered by the user.

[1461] Output: Data that identifies the user's interests and preferences as a result of the analysis. For example, if the keyword "travel" is included, it is determined that the user is interested in travel.

[1462] Step 3:

[1463] Generate prompt statement

[1464] Based on the analysis results, the server creates a prompt sentence for the generative AI model to generate a message.

[1465] Input: User interests and preferences identified as a result of the analysis.

[1466] Output: A prompt to send to the generative AI model. For example, "The user seems to have recently become interested in traveling. Please generate a message with suggestions for new travel destinations."

[1467] Step 4:

[1468] Message Generation

[1469] The generative AI model generates a message that matches the user's interests and preferences based on the prompt text received.

[1470] Input: The prompt text sent by the server.

[1471] Output: A generated message to be presented to the user. For example, "Hello! I heard you've been interested in traveling lately. How about a recommended destination, Tuscany, Italy? You'll enjoy beautiful scenery and delicious food."

[1472] Step 5:

[1473] Providing a message

[1474] The server sends the generated message to the user's terminal.

[1475] Input: The generated message from the generative AI model.

[1476] Output: The message sent to the user's terminal.

[1477] Step 6:

[1478] Collecting user responses

[1479] Users can react to the generated messages, for example by rating the messages and providing feedback.

[1480] Input: User response (ratings and feedback).

[1481] Output: User responses sent from the device to the server.

[1482] Step 7:

[1483] Training generative AI models

[1484] The server collects user responses and provides them as feedback to the generative AI model.

[1485] Input: User response data.

[1486] Output: The generative AI model learns and reflects this in future message generation. For example, if a user rates the "Tuscany" suggestion highly, the model will likely suggest similar travel destinations next time.

[1487] (Application example 1)

[1488] 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."

[1489] Conventional content recommendation systems have difficulty fully understanding a user's interests and preferences and recommending appropriate content. It is also difficult to generate messages that address emotions and desires that users cannot verbalize. Furthermore, they lack the ability to continuously adjust messages based on user responses, making it difficult to improve the user experience. To solve these issues, a system is needed that can deeply understand a user's interests and preferences and recommend new content based on that understanding.

[1490] 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.

[1491] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, and a means for understanding the user's interests and preferences and recommending new content based on those. This makes it possible to recommend appropriate content based on the user's interests and preferences, and to generate messages that respond to emotions and desires that the user cannot verbalize. Furthermore, by continuously adjusting the messages based on the user's reactions, the user experience can be improved.

[1492] "Generative AI methods" are artificial intelligence methods that use natural language processing technology to understand a user's interests and preferences and generate messages based on them.

[1493] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[1494] The "means for understanding the user's interests and preferences" refers to a means for analyzing the user's past behavioral data and input information to understand the user's interests and preferences.

[1495] The "means for recommending new content" is a means for selecting and recommending new content that the user does not yet know based on the user's interests and preferences.

[1496] "Means for adjusting messages based on user responses" refers to a means by which the generative AI means learns feedback from users and reflects it in future message generation.

[1497] A system for implementing this invention includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for understanding the user's interests and preferences and recommending new content based on those interests and preferences. Specific embodiments are described below.

[1498] The server uses OpenAI's API as a generative AI means. It analyzes the user's past behavioral data and input information to understand the user's interests and preferences. This analysis is performed using programming languages ​​such as Python. To recommend new content based on the user's interests and preferences, the generative AI means generates prompts and generates messages using OpenAI's API.

[1499] As a concrete example, if the user's interests are "sci-fi," "action," and "comedy," the generated recommendation message would look like this:

[1500] "Based on your interests, we recommend the new movie 'Interstellar,' the music 'Action Movie Soundtracks,' and the article 'Latest Comedy Movie Reviews.'"

[1501] Example prompt sentence:

[1502] "The user's interests and preferences are ['sci-fi', 'action', 'comedy']. Recommend new movies, music, and articles based on this."

[1503] To tailor messages based on user responses, the generative AI method learns from user feedback and incorporates it into future message generation. This feedback is learned using machine learning algorithms.

[1504] This system makes it possible to recommend appropriate content based on the user's interests and preferences, and to generate messages that respond to emotions and desires that the user cannot verbalize. Furthermore, by continuously adjusting messages based on the user's reactions, it is possible to improve the user experience.

[1505] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1506] Step 1:

[1507] The server obtains the user's interests and preferences.

[1508] Input: User ID

[1509] Output: User's interests and preferences (e.g. "sci-fi", "action", "comedy")

[1510] Specific operation: The server retrieves interest and preference data corresponding to the user ID from the database. This data is based on the user's past behavior history and input information.

[1511] Step 2:

[1512] The server generates prompts based on the user's interests and preferences.

[1513] Input: User interests and preferences

[1514] Output: Prompt (e.g., "The user's interests and preferences are ['sci-fi', 'action', 'comedy']. Based on this, please recommend new movies, music, and articles.")

[1515] Specific operation: Based on the acquired user interests and preferences, the server generates a prompt sentence to be input into the generative AI model.

[1516] Step 3:

[1517] The server generates recommendation messages using a generative AI model.

[1518] Input: prompt statement

[1519] Output: Recommendation message (e.g., "Based on your interests, we recommend the new movie 'Interstellar,' the music 'Action Movie Soundtracks,' and the article 'Latest Comedy Movie Reviews.'")

[1520] Specific operation: The server sends the generated prompt sentence to OpenAI's API and generates a recommendation message using the generative AI model.

[1521] Step 4:

[1522] The server sends the generated recommendation message to the user.

[1523] Input: Recommendation message, user ID

[1524] Output: Recommendation message displayed on the user's device

[1525] Specific operation: The server sends the generated recommendation message to the user's terminal, allowing the user to receive the message.

[1526] Step 5:

[1527] The user sends a response to the recommendation message to the server.

[1528] Input: User response (e.g. clicks, feedback comments)

[1529] Output: User response data sent to the server

[1530] Specific operation: The user clicks or makes a feedback comment on the recommendation message, and sends the reaction data to the server.

[1531] Step 6:

[1532] The server learns the user's reaction data and reflects it in subsequent message generation.

[1533] Input: User response data

[1534] Output: An updated generative AI model

[1535] Specific operation: The server analyzes the user's reaction data and adjusts the parameters of the generation AI model to reflect the results in future message generation. This enables the server to generate messages that are more suited to the user's interests and preferences.

[1536] Example 2

[1537] 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."

[1538] Conventional generative AI systems have had difficulty accurately grasping user interests and effectively recommending new content. Furthermore, they lacked a means to appropriately provide feedback on user reactions and optimize the generative AI model, making it difficult to increase user satisfaction.

[1539] 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.

[1540] In this invention, the server includes means for collecting user behavior data, means for analyzing the collected data to determine the user's interests, a generation AI means for generating messages recommending new content based on the user's interests, means for collecting user responses to the generated messages, and means for providing feedback to the generation AI means based on the collected user responses to allow it to learn. This makes it possible to accurately grasp the user's interests and recommend appropriate content. Furthermore, optimizing the generation AI model based on the user's responses can increase user satisfaction.

[1541] "User behavior data" refers to information such as the operations and selections a user makes within the system, playback history, and ratings.

[1542] "Means for collection" refers to functions and devices for acquiring and storing user behavioral data.

[1543] "Means for analyzing and determining user interests" refers to functions and devices for analyzing collected user behavior data and identifying genres and content in which the user is interested.

[1544] "Generative AI means" refers to artificial intelligence models or algorithms that generate messages recommending new content based on a user's interests.

[1545] "Means for collecting user responses" refers to a function or device for recording and storing user responses (e.g., positive and negative responses) to generated messages.

[1546] "Means of feedback and learning" refers to functions and devices that adjust the generative AI means based on collected user responses and retrain the model.

[1547] "New content" refers to information about music, artists, videos, articles, and other things that users may not yet know about or that may be of interest to them.

[1548] "Message" refers to text or notifications generated to recommend new content to users.

[1549] MODE FOR CARRYING OUT THE INVENTION

[1550] The present invention is a system that collects and analyzes user behavior data to determine the user's interests and generates messages recommending new content. A specific embodiment of this system will be described below.

[1551] Collecting user behavior data

[1552] The server uses the APIs of music streaming services and other data sources to collect user behavior data. For example, it uses APIs of Spotify and Apple Music to obtain user playback history and rating data. This allows it to understand what kind of music users listen to and which artists they like.

[1553] User interest analysis

[1554] The server analyzes the collected behavioral data to determine the user's interests. Specifically, it uses a data analysis algorithm to identify the genres and artists the user frequently listens to. Based on the analysis results, the server understands the user's musical preferences.

[1555] Generating recommendation messages for new content

[1556] The server uses a generative AI model to generate messages that recommend new content based on the user's interests. The generative AI model receives prompt sentences that reflect the user's interests as input and generates messages based on them.

[1557] As a concrete example, the following prompt sentence is input to the generative AI model:

[1558] The genre that the user has been listening to recently is "Rock," and they especially like "Queen." Please generate a message that introduces new rock artists and songs that the user has not yet heard of.

[1559] Based on this prompt, the generative AI model generates a message like this:

[1560] "Hi! I see you've been listening to a lot of Queen lately. If you haven't heard them yet, I recommend The Black Keys. I think you'll love 'Lonely Boy'!"

[1561] Collecting user responses

[1562] The device collects user reactions to the generated messages. When a user takes an action such as "Like" or "Skip" on a message, the device records the reaction data and sends it to the server.

[1563] Feedback and training for generative AI models

[1564] The server feeds back the generated AI model based on the collected user reaction data and trains it. This allows the generated AI model to optimize recommendation messages to match the user's preferences. By increasing the frequency of messages with a high percentage of positive reactions and decreasing the frequency of messages with a high percentage of negative reactions, user satisfaction can be increased.

[1565] In this way, a system is realized that collects and analyzes user behavioral data, uses a generative AI model to generate messages recommending new content, and optimizes the model based on feedback from users.

[1566] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1567] Step 1:

[1568] The server collects user behavior data. Specifically, it uses the music streaming service's API to obtain the user's playback history and rating data. It uses the user's authentication information and API key as input and obtains the user's playback history data as output. The server stores this data in a database.

[1569] Step 2:

[1570] The server analyzes the collected data to determine the user's interests. Specifically, it uses a data analysis algorithm to identify the genres and artists the user frequently listens to. As input, it uses the playback history data stored in the database, and as output, it obtains an analysis result indicating the user's interests. The server uses this analysis result in the next step.

[1571] Step 3:

[1572] The server uses the generative AI model to generate a message recommending new content based on the user's interests. Specifically, it generates a prompt sentence that reflects the user's interests and inputs it into the generative AI model. The prompt sentence created based on the analysis results is used as input, and a recommendation message is obtained as output. The server then sends this message to the user.

[1573] Step 4:

[1574] The device collects the user's reaction to the generated message. Specifically, it records whether the user performed an action such as "like" or "skip" on the message. The device uses the user's operation information as input and obtains the user's reaction data as output. The device then transmits this reaction data to the server.

[1575] Step 5:

[1576] The server feeds back the generative AI model based on the collected user reaction data and trains it. Specifically, it classifies reactions into positive and negative and provides feedback to the generative AI model. The user reaction data is used as input, and an optimized generative AI model is obtained as output. The server uses this optimized model to generate the next message.

[1577] (Application example 2)

[1578] 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."

[1579] Conventional generative AI systems have had difficulty recommending appropriate content based on a user's interests and preferences, which has led to the issue of being unable to increase user satisfaction. Additionally, they lacked a mechanism for appropriately utilizing user responses as feedback and reflecting them in subsequent recommendations, making it difficult to continuously provide content that matches the user's interests.

[1580] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for recording user responses and adjusting the frequency of message generation based on the responses. This makes it possible to recommend content based on the user's interests and preferences, thereby increasing user satisfaction. Furthermore, by appropriately utilizing the user's responses as feedback and reflecting them in the next recommendation, it becomes possible to continuously provide content that matches the user's interests.

[1581] "Generative AI means" refers to artificial intelligence technology that generates messages based on a user's interests and preferences and sends them to the user.

[1582] "Means for guiding users to services" refers to a mechanism for guiding users to specific services or content based on messages sent by generative AI means.

[1583] "Means for recording user responses" refers to technology that records as data the responses that users show to messages sent by generative AI means.

[1584] "Means for adjusting the frequency of message generation" refers to a mechanism by which the generative AI means dynamically changes the frequency at which messages are generated based on recorded user responses.

[1585] "Content" refers to the information and entertainment that users consume, such as music, videos, and articles.

[1586] "Recommendation" refers to the act of suggesting new content based on a user's interests and preferences.

[1587] A system for implementing this invention includes a generation AI means for sending messages to users, a means for directing users to services based on the messages sent by the generation AI means, and a means for recording user responses and adjusting the frequency of message generation based on the responses.

[1588] The server uses a generative AI means to generate messages based on the user's interests and preferences and sends them to the user's device. This generative AI means is implemented using a programming language such as Python and recommends appropriate content by analyzing the user's past behavioral and reaction data.

[1589] When a user receives a message and responds positively or negatively to it, the server records the response. This record is saved in a data format such as JSON. The user's response data is used as feedback when generating the next message, and the generative AI method adjusts the frequency of message generation.

[1590] As a concrete example, consider a user using a music streaming service. A generative AI tool generates a message recommending new artists or songs that the user has not yet heard of and sends it to the user's smartphone. If the user responds positively to the message, the server records the response and increases the frequency of similar recommendations next time. On the other hand, if the user responds negatively, the server decreases the frequency of similar recommendations.

[1591] Examples of prompts include:

[1592] "When a user is recommended 'Artist A - Song 1', record the action taken if they respond positively."

[1593] In this way, the server can continually provide content based on the user's interests and preferences, increasing user satisfaction.

[1594] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1595] Step 1:

[1596] The server collects users' past behavioral and reaction data, including the content they have viewed, the content they have rated, their search history, etc. The input is the user's behavioral data, and the output is a list of the collected data.

[1597] Step 2:

[1598] The server analyzes the collected data to identify the user's interests and preferences. Specifically, it uses machine learning algorithms to extract patterns of content that the user prefers. The input is a list of collected data, and the output is a profile that indicates the user's interests and preferences.

[1599] Step 3:

[1600] The server uses generative AI tools to generate messages based on the user's profile, including recommendations for new content the user does not yet know. The input is the user's profile, and the output is the generated message.

[1601] Step 4:

[1602] The server sends the generated message to the user's terminal. The input is the generated message, and the output is the message displayed on the user's terminal.

[1603] Step 5:

[1604] The user responds positively or negatively to the message displayed on the terminal. The input is the user's reaction, and the output is the reaction data.

[1605] Step 6:

[1606] The server records the user's reaction data. The input is the user's reaction data, and the output is a list of the recorded reaction data.

[1607] Step 7:

[1608] The server provides feedback based on the recorded reaction data when generating the next message. Specifically, if there are many positive reactions, it increases the number of similar recommendations, and if there are many negative reactions, it decreases the number of recommendations. The input is a list of recorded reaction data, and the output is the adjusted message generation frequency.

[1609] Example 3

[1610] 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."

[1611] In conventional systems, when users wanted to find a specific product, they had to search for it themselves, which was time-consuming and labor-intensive. It was also difficult to provide appropriate information for content that users could not verbalize or were completely unfamiliar with. Furthermore, the system lacked the ability to adjust messages based on user responses, leaving a need for an improved user experience.

[1612] 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.

[1613] In this invention, the server includes a generative AI means for sending messages to users, a means for guiding users to services based on the messages sent by the generative AI means, a means for collecting user behavioral data, a means for analyzing the collected behavioral data to determine the user's intention, a means for generating a prompt sentence based on the user's intention, a means for inputting the generated prompt sentence into a generative AI model to generate a message, and a means for displaying the generated message on the user's device. This eliminates the need for users to search for information themselves and provides appropriate information even for content that cannot be verbalized or is completely unknown. Furthermore, adjusting the message based on the user's response can improve the user experience.

[1614] "Generative AI means" refers to means that use artificial intelligence technology to generate and send messages to users.

[1615] "Means for guiding users to services" refers to means for guiding users to specific services or content based on messages sent by generative AI means.

[1616] The "means for collecting behavioral data" refers to a means for collecting behavioral data such as a user's search history and browsing history.

[1617] "Means for analyzing behavioral data to determine user intent" refers to means for analyzing collected behavioral data to determine what the user is looking for and what their intent is.

[1618] The "means for generating a prompt sentence" is a means for generating a prompt sentence to be input to a generative AI model based on the user's intention.

[1619] "Means for inputting into a generative AI model to generate a message" refers to means for inputting the generated prompt sentence into a generative AI model to generate a message introducing related information or other products in the same category.

[1620] The "means for displaying a message on a user's terminal" refers to a means for transmitting the generated message to the user's terminal and displaying it.

[1621] This invention relates to a system that, when it determines that a user is looking for a specific product, generates a message introducing information related to that product or other products in the same category. The system includes generative AI means, means for collecting user behavioral data, means for analyzing the behavioral data to determine the user's intention, means for generating a prompt sentence, means for inputting the prompt sentence into a generative AI model to generate a message, and means for displaying the generated message on the user's device.

[1622] The server collects the user's search history and browsing history using a database management system (e.g., MySQL). The collected data is analyzed using machine learning algorithms (e.g., random forests and support vector machines) to determine whether the user is looking for a specific product. For example, if it determines that the user is looking for a "smartphone," the server generates a prompt like the following:

[1623] "User is searching for a smartphone. Generate a message that provides relevant information and introduces other smartphones in the same category."

[1624] The generated prompt is then fed into a generative AI model (e.g., OpenAI's GPT-4) to generate a message introducing related information or other products in the same category. The generated message is then sent to the user's device and displayed on a web page or application interface.

[1625] For example, if the user is searching for a "smartphone," the server generates a message like this:

[1626] "Here's some information related to the phone you're looking for. We also recommend the following phones: iPhone 13, Samsung Galaxy S21, Google Pixel 6."

[1627] This system saves users the trouble of searching for information themselves and provides appropriate information for content that they cannot verbalize or are completely unfamiliar with. Furthermore, by adjusting messages based on the user's reactions, the user experience can be improved. The flow of the identification process in Example 3 will be described with reference to FIG. 15.

[1628] Step 1:

[1629] The server collects user behavior data. Specifically, it obtains information about the search keywords and products viewed by users on websites and applications from a database management system (e.g., MySQL). The input is the user's search history and browsing history, and the output is the collected behavioral data.

[1630] Step 2:

[1631] The server analyzes the collected behavioral data to determine the user's intent. Specifically, it uses machine learning algorithms (e.g., random forests and support vector machines) to analyze the user's behavioral patterns. The input is the collected behavioral data, and the output is a determination of whether the user is looking for a specific product.

[1632] Step 3:

[1633] The server generates a prompt based on the user's intent. For example, if it determines that the user is looking for a "smartphone," it generates the following prompt: "The user is looking for a smartphone. Please generate a message introducing related information and other smartphones in the same category." The input is the result of the user's intent determination, and the output is the generated prompt.

[1634] Step 4:

[1635] The server inputs the generated prompt sentence into a generative AI model to generate a message. Specifically, the prompt sentence is input into a generative AI model (e.g., OpenAI's GPT-4) to generate a message introducing related information or other products in the same category. The input is the generated prompt sentence, and the output is the generated message.

[1636] Step 5:

[1637] The server displays the generated message on the user's terminal. Specifically, it sends the generated message to the user's terminal and displays it on a web page or application interface. The input is the generated message, and the output is the message displayed on the user's terminal.

[1638] Step 6:

[1639] The user checks the displayed message and clicks to check detailed information as necessary. Specifically, when the user clicks on the displayed message, detailed information is displayed. The input is the user's click operation, and the output is the display of detailed information.

[1640] (Application example 3)

[1641] 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."

[1642] On conventional online shopping sites, when users search for a specific product, they have to search for related products or other products in the same category, which is time-consuming. It is also difficult for users to find the right product for needs they cannot verbalize or content they are completely unfamiliar with. Furthermore, there is a lack of functionality to adjust messages based on user responses, leaving a need for an improved user experience.

[1643] 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.

[1644] In this invention, the server includes a generative AI means for sending messages to users, a means for directing users to services based on the messages sent by the generative AI means, and a means for analyzing the user's search query and recommending related products or other products in the same category. This eliminates the need for the user to perform a search themselves and automatically recommends related products or other products in the same category. Furthermore, it is possible to recommend appropriate products for needs that the user cannot verbalize or content that the user is completely unfamiliar with, and by adjusting the messages based on the user's responses, the user experience can be improved.

[1645] "Generative AI methods" are artificial intelligence technologies that automatically generate related information and products based on user behavior and search queries and send them as messages.

[1646] "Means for guiding users to services" is a function for guiding users to specific services or product pages based on messages sent by generative AI means.

[1647] "Means for analyzing search queries" refers to technology that analyzes search keywords and phrases entered by users to understand their intent and relevance.

[1648] "Means for recommending related products" is a function that automatically selects and recommends products that may interest a user based on the user's search queries and behavioral history.

[1649] The "means for recommending other products in the same category" is a function for automatically selecting and recommending other products that belong to the same category as the product searched for by the user.

[1650] The system for implementing the present invention has a function of analyzing a user's search query and recommending related products or other products in the same category. A specific embodiment of this system is described below.

[1651] System Configuration

[1652] The system consists of the following main components:

[1653] 1. Generative AI methods: Artificial intelligence technology that automatically generates relevant information and products based on user behavior and search queries and sends them as messages.

[1654] 2. Means of guiding users to services: A function to guide users to specific services or product pages based on messages sent by generative AI means.

[1655] 3. Search query analysis: Techniques for analyzing search keywords and phrases entered by users to understand their intent and relevance.

[1656] 4. A means of recommending related products: A feature that automatically selects and recommends products that may be of interest to users based on their search queries and behavioral history.

[1657] 5. A means of recommending other products in the same category: A function that automatically selects and recommends other products that belong to the same category as the product searched by the user.

[1658] Hardware and software used

[1659] Hardware: Smartphones, servers

[1660] Software: Python, OpenAI API

[1661] Data processing and calculation

[1662] The server receives the search query entered by the user and inputs it as a prompt into the generative AI model. The generative AI model generates a message recommending related products or other products in the same category based on the prompt. The generated message is sent to the user's smartphone, where the user can view it.

[1663] Specific examples

[1664] If a user searches for "running shoes" on their smartphone, the server receives this search query and inputs the following prompt sentence into the generative AI model:

[1665] Example prompt sentence:

[1666] User is searching for 'running shoes'. Recommend related products or other products in the same category.

[1667] Based on this prompt, the generative AI model generates a recommendation message like the following:

[1668] An example of the message generated:

[1669] Looking for running shoes? We also recommend:

[1670] Running wear

[1671] Fitness tracker

[1672] Running socks

[1673] Sports Watch

[1674] This message is sent to the user's smartphone, allowing them to easily find related products or other products in the same category, saving the user the trouble of searching for them themselves and improving convenience.

[1675] The flow of the specific processing in Application Example 3 will be described with reference to FIG.

[1676] Step 1:

[1677] The user searches for a specific product on their device (smartphone).

[1678] Input: The search query entered by the user (e.g., "running shoes")

[1679] Output: The search query is sent to the server.

[1680] Specific action: A user types "running shoes" into the search bar on their smartphone and presses the search button.

[1681] Step 2:

[1682] The server receives and parses the search query.

[1683] Input: The search query submitted by the user

[1684] Output: Search query analysis results (user intent and relevance)

[1685] Specific operation: The server receives the search query "running shoes" and analyzes it using natural language processing techniques.

[1686] Step 3:

[1687] The server generates a prompt based on the analysis results.

[1688] Input: Analysis result of search query

[1689] Output: Generated prompt statement

[1690] Specific behavior: The server generates a prompt sentence such as "The user is looking for 'running shoes'. Please recommend related products or other products in the same category."

[1691] Step 4:

[1692] The server inputs a prompt sentence into the generative AI model, which generates a message recommending related products or other products in the same category.

[1693] Input: Generated prompt statement

[1694] Output: Recommendation message

[1695] Specific operation: The server uses the OpenAI API to input a prompt sentence into the generative AI model and generate a recommendation message.

[1696] Step 5:

[1697] The server transmits the generated recommendation message to the user's terminal.

[1698] Input: Recommendation message

[1699] Output: A recommendation message that is displayed on the user's device.

[1700] Specific operation: The server sends the generated recommendation message to the user's smartphone, allowing the user to view the message.

[1701] Step 6:

[1702] The user views the recommendation messages and selects the product of interest.

[1703] Input: Recommendation message

[1704] Output: The product page selected by the user

[1705] Specific behavior: The user checks the recommendation message on their smartphone and taps on the product they are interested in to go to the detail page.

[1706] Step 7:

[1707] The server records the user's responses and accumulates the data to improve the accuracy of recommendations in future.

[1708] Input: User selected data

[1709] Output: Accumulated user behavior data

[1710] Specific operation: The server saves the user's selection data in a database and uses it to improve the accuracy of recommendations in the future.

[1711] 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.

[1712] "Example 1"

[1713] In one embodiment of the present invention, a system includes a generative AI means including an emotion engine that recognizes a user's emotion. The emotion engine infers the user's emotion from the user's facial expression, tone of voice, message context, etc. For example, if the user expresses joy, the emotion engine conveys that information to the generative AI means, which then generates a joy-related message based on that information.

[1714] "Example 2"

[1715] Another embodiment of the present invention is a system in which an emotion engine generates a message based on a user's emotion. Specifically, when a user feels sad, the emotion engine transmits that information to a generative AI means, which then generates a comforting message based on that information.

[1716] "Example 3"

[1717] Furthermore, one embodiment of the present invention is a system in which an emotion engine adjusts messages in response to changes in the user's emotions. For example, if the user initially expresses joy but suddenly expresses sadness, the emotion engine conveys this information to the generative AI means, which then adjusts the message based on this information and generates a comforting message.

[1718] The processing flow of each embodiment will be described below.

[1719] "Example 1"

[1720] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[1721] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[1722] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotions.

[1723] "Example 2"

[1724] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[1725] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[1726] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotions. In particular, if the user feels sad, the generative AI means generates a comforting message.

[1727] "Example 3"

[1728] Step 1: The user provides some input to the system. This input can take many forms, including text messages, voice, and images.

[1729] Step 2: The emotion engine infers emotions from the user's input using techniques such as natural language processing, speech analysis, and image analysis.

[1730] Step 3: The generative AI means generates a message based on the information from the emotion engine. This message changes depending on the user's emotion. In particular, if the user's emotion changes, the generative AI means adjusts the message accordingly.

[1731] Example 1

[1732] 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."

[1733] Conventional generative AI systems have had difficulty accurately understanding users' interests and preferences and generating appropriate messages based on them. Furthermore, they lacked the ability to capture users' reactions in real time and adjust messages based on their emotions, making it difficult to improve the user experience.

[1734] The identification process by the identification processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means. In this invention, the server includes generative artificial intelligence means, means for guiding a user to a service based on a message transmitted by the generative artificial intelligence means, means for inputting information on the user's interests and preferences, means for analyzing the input information, means for capturing the user's reactions, means for analyzing the captured reactions and recognizing the user's emotions, and means for adjusting the message based on the recognized emotions. This makes it possible to accurately understand the user's interests and preferences, capture the user's reactions in real time to recognize their emotions, and adjust the message based on the emotions.

[1735] "Generative AI means" refers to AI technology that generates messages based on a user's interests and preferences.

[1736] "Means for guiding users to a service" refers to a function for guiding users to a specific service or action based on a message sent by the generative artificial intelligence means.

[1737] "Means for inputting information about user interests and preferences" refers to an interface or method by which a user provides information about their interests and preferences to the system.

[1738] "Means for analyzing input information" refers to the technology or algorithms used to analyze the information entered by the user and understand its content.

[1739] "Means for capturing user responses" refers to technologies or devices that record a user's response (e.g., facial expression, tone of voice, text) to a generated message.

[1740] "Means for recognizing a user's emotions by analyzing captured reactions" refers to technology or algorithms for analyzing a user's captured reactions and inferring the user's emotions from those reactions.

[1741] "Means for adjusting a message based on recognized emotions" refers to a function for appropriately changing or adjusting a message generated by a generative artificial intelligence means based on the recognized emotions of a user.

[1742] MODE FOR CARRYING OUT THE INVENTION

[1743] The present invention is a system that uses generative artificial intelligence to understand a user's interests and preferences and generate messages based on them. The system includes the ability to capture the user's reactions, recognize their emotions, and adjust the messages accordingly.

[1744] Hardware and software used

[1745] Hardware: Servers, devices (PCs, smartphones, etc.)

[1746] Software: Generative AI models (e.g., GPT-4), emotion engines, natural language processing libraries (e.g., spaCy, NLTK)

[1747] System Overview

[1748] 1. Input of user interests and preferences

[1749] User: A user uses a terminal to access the system and enter information about their interests and preferences, for example, "Where would you like to go on your next vacation?"

[1750] 2. Analysis of input data

[1751] Terminal: Sends the information entered by the user to the server.

[1752] Server: The server analyzes the received data using natural language processing libraries. Specifically, it extracts keywords and context from the input text and understands the user's interests and preferences.

[1753] 3. Message Creation

[1754] Server: Based on the analysis results, the server sends a prompt to the generative AI model. An example of a prompt is, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[1755] Generative AI: Generative AI receives a prompt and generates an appropriate message, such as, "Hawaii is a recommended destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[1756] 4. Capturing User Reactions

[1757] Device: The user receives the generated message. The device captures the user's reactions (facial expressions, tone of voice, message context, etc.). For example, if the user shows a happy expression after reading the message, the device captures that expression with a camera.

[1758] 5. Emotional Recognition

[1759] Terminal: Sends captured data to the server.

[1760] Server: The server uses an emotion engine to analyze the captured data and infer the user's emotions, for example, detecting the emotion of joy from the user's facial expression.

[1761] 6. Tailor your message

[1762] Server: Based on the information from the emotion engine, sends a new prompt to the generative AI. An example of a prompt is, "The user is expressing joy. Please provide additional information related to that emotion."

[1763] Generative AI: Generative AI receives new prompts and generates additional messages, such as "I'll also introduce some recommended tourist spots in Hawaii!"

[1764] Specific examples

[1765] User: A user interested in traveling comes into the system and types, "Where would you like to go on your next vacation?"

[1766] Terminal: Sends user input to the server.

[1767] Server: Using a natural language processing library, extract the keyword "travel destination" and understand the user's interest as "travel."

[1768] Server: Sends the generative AI a prompt saying, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[1769] Generative AI: Generates a message like, "Hawaii is a great destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[1770] Device: The user receives this message and shows a look of delight.

[1771] Device: The camera captures the user's facial expressions and sends the data to the server.

[1772] Server: Uses an emotion engine to detect the emotion of joy from the user's facial expression.

[1773] Server: Sends the prompt to the generative AI: "The user is expressing joy. Please provide additional information related to that emotion."

[1774] Generative AI: Generates an additional message such as, "We'll also introduce some recommended tourist spots in Hawaii!"

[1775] The above is an embodiment of the present invention.

[1776] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1777] Step 1:

[1778] User: A user uses a terminal to access the system and enter information about their interests and preferences, for example, "Where would you like to go on your next vacation?"

[1779] Input: Text information about the user's interests and preferences

[1780] Output: User input data

[1781] Step 2:

[1782] Terminal: Sends the information entered by the user to the server.

[1783] Input: User-entered data

[1784] Output: User input data sent to the server

[1785] Step 3:

[1786] Server: The server analyzes the received data using a natural language processing library (e.g., spaCy, NLTK). Specifically, it extracts keywords and context from the input text and understands the user's interests and preferences.

[1787] Input: User-entered data

[1788] Data processing: Text analysis using natural language processing libraries

[1789] Output: Analysis of user interests and preferences

[1790] Step 4:

[1791] Server: Based on the analysis results, the server sends a prompt to a generative AI model (e.g., GPT-4). An example of a prompt is, "Please suggest recommended travel destinations for the user's next vacation based on their interests and preferences."

[1792] Input: Analysis results about user interests and preferences

[1793] Data processing: Prompt generation

[1794] Output: Prompt for the generative AI model

[1795] Step 5:

[1796] Generative AI: Generative AI receives a prompt and generates an appropriate message, such as, "Hawaii is a recommended destination for your next vacation. You can enjoy beautiful beaches and abundant nature."

[1797] Input: prompt statement

[1798] Data processing: Message generation

[1799] Output: The generated message

[1800] Step 6:

[1801] Device: The user receives the generated message. The device captures the user's reactions (facial expressions, tone of voice, message context, etc.). For example, if the user shows a happy expression after reading the message, the device captures that expression with a camera.

[1802] Input: The generated message

[1803] Output: User response data

[1804] Step 7:

[1805] Terminal: Sends captured data to the server.

[1806] Input: User response data

[1807] Output: User response data sent to the server

[1808] Step 8:

[1809] Server: The server uses an emotion engine to analyze the captured data and infer the user's emotions, for example, detecting the emotion of joy from the user's facial expression.

[1810] Input: User response data

[1811] Data processing: Emotion analysis using an emotion engine

[1812] Output: Analysis results on user sentiment

[1813] Step 9:

[1814] Server: Based on the information from the emotion engine, sends a new prompt to the generative AI. An example of a prompt is, "The user is expressing joy. Please provide additional information related to that emotion."

[1815] Input: Analysis results on user sentiment

[1816] Data processing: generating new prompts

[1817] Output: A new prompt for the generative AI model

[1818] Step 10:

[1819] Generative AI: Generative AI receives new prompts and generates additional messages, such as "I'll also introduce some recommended tourist spots in Hawaii!"

[1820] Input: new prompt text

[1821] Data processing: generating additional messages

[1822] Output: Additional messages generated

[1823] (Application example 1)

[1824] 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."

[1825] Conventional content delivery services recommend content based on users' interests and preferences, but they lack personalized recommendations that take into account the user's current emotions. Furthermore, it is difficult to generate messages that respond to emotions and desires that users cannot verbalize. This limits the improvement of the user experience.

[1826] 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: a generative AI means for sending a message to a user; a means for guiding the user to a service based on the message sent by the generative AI means; an emotion recognition means for recognizing emotions by analyzing the user's facial expressions and tone of voice; a preference understanding means for understanding the user's interests and preferences by analyzing the user's past viewing history and search history; and a content recommendation means for recommending personalized content based on the emotion recognition means and preference understanding means. This enables personalized content recommendation based on the user's current emotions and past preferences, improving the user experience.

[1827] "Generative AI methods" are artificial intelligence methods that understand a user's interests and preferences and generate messages based on them.

[1828] "Means for guiding users to services" refers to means for guiding users to specific services or actions based on messages sent by generative AI means.

[1829] The "emotion recognition means" is a means for recognizing the user's emotions by analyzing the user's facial expressions and tone of voice.

[1830] The "preference understanding means" is a means for understanding the user's interests and preferences by analyzing the user's past viewing history and search history.

[1831] The "content recommendation means" is a means for recommending personalized content to a user based on the emotion recognition means and the preference understanding means.

[1832] A system for implementing the present invention includes the following means for recommending personalized content based on a user's emotions and preferences.

[1833] System Configuration

[1834] 1. Generative AI means:

[1835] The server uses generative AI to understand the user's interests and preferences, and generates messages based on them. The generated messages address feelings and desires that the user cannot verbalize, or introduce new content that the user is completely unfamiliar with.

[1836] 2. Means of directing you to the service:

[1837] The server guides the user to specific services or actions based on the messages sent by the generative AI means, allowing the user to access the right content or services at the right time.

[1838] 3. Emotion recognition means:

[1839] The device uses a camera and microphone to recognize emotions by analyzing the user's facial expressions and tone of voice. The emotion recognition engine analyzes this data in real time to estimate the user's current emotion.

[1840] 4. Means of understanding preferences:

[1841] The server analyzes the user's past viewing and search history to understand their interests and preferences, and uses natural language processing technology to extract these from the text entered by the user.

[1842] 5. Content recommendation methods:

[1843] The server recommends personalized content to users based on emotion recognition and preference understanding. The recommendation algorithm learns user responses and is continually adjusted.

[1844] Hardware and software used

[1845] Hardware:

[1846] Devices such as smartphones and tablets

[1847] Camera and microphone

[1848] software:

[1849] Emotion recognition engines (e.g., ...

Claims

1. A generating AI means for sending a message to a user; A means for guiding the user to a service based on a message transmitted by the generating AI means; A means for collecting behavioral data of the user; means for analyzing collected behavioral data to determine the intent of the user; a means for generating a prompt sentence for instructing to generate a message introducing at least one of related information and other products in the same category based on the user's intention; A means for inputting the generated prompt sentence into a generative AI model to generate a message introducing at least one of related information and other products in the same category; means for displaying the generated message on a terminal of the user; A system including:

2. The system of claim 1 , wherein the generative AI means generates messages for content that the user cannot verbalize or is completely unaware of.

3. The system of claim 1 , wherein the generative AI means adjusts messages based on the user's responses.

Images