System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting data when a user uses a generative artificial intelligence; means for transmitting the collected data to a server; means for analyzing the transmitted data; means for enhancing model learning of the generative artificial intelligence based on an analysis result; and means for delivering an updated generative artificial intelligence model to a terminal.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Conventional generative AI still has issues with accuracy and sophistication, and is unable to fully meet the diverse needs of users. Furthermore, improving the learning model of generative AI requires a lot of time and effort, and more efficient data collection and optimization of the learning process are required. It is necessary to solve these issues and provide users with highly accurate and advanced generative AI services. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for collecting data when a user uses a generative AI, a means for transmitting the collected data to a server, a means for analyzing the transmitted data, a means for enhancing the generative AI's model learning based on the analysis results, a means for distributing an updated generative AI model to a terminal, and a means for providing a reward to the user. This system efficiently collects 5W1H data (when, where, who, what, why, and how) and enhances the generative AI's learning based on the analysis results, making it possible to provide a highly accurate generative AI that meets the user's needs.

[0006] "User" refers to an individual or corporation that uses the Generative Artificial Intelligence.

[0007] "Generative AI" refers to artificial intelligence technology for generating data in various formats, such as natural language generation and image generation.

[0008] "Data" refers to information generated when a user uses generative artificial intelligence, including information related to the 5W1H (when, where, who, what, why, and how).

[0009] "Means" refers to apparatus, software, or a combination thereof for accomplishing a particular purpose.

[0010] "Server" refers to a computer system that communicates with terminals via a network and manages, processes, and analyzes data.

[0011] "Model learning" refers to the process of training generative AI algorithms based on collected data to improve their accuracy and performance.

[0012] "Analysis" refers to the process of compiling, classifying, and performing pattern recognition on collected data to extract meaningful information.

[0013] "Distribution" refers to the act of sending an updated generative artificial intelligence model from a server to a terminal. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0035] This invention relates to a system that provides more accurate generative AI services by collecting data on the use of generative AI and enhancing the learning of AI models through analysis. Below, we will explain how this system is implemented by the server, terminal, and user.

[0036] Data collection:

[0037] When a user begins using the generation AI, the device automatically detects the user's actions and begins recording data. The device collects data such as the start and end times of use (when), IP address and GPS information (where), user ID and device ID (who), the type of generation request and specific usage (what), purpose information entered by the user (why), and specific instructions and conditions (how).

[0038] Data transmission:

[0039] The collected data is encrypted by the device and sent using a secure communication protocol to a server, which verifies the received data and stores it in a database.

[0040] Data Analysis:

[0041] The server analyzes the received data using aggregation, classification, and pattern recognition algorithms to identify usage frequency, regional needs, and usage trends by time of day, resulting in a detailed analysis of the use of the Generative AI.

[0042] Strengthening model training:

[0043] The server uses data analysis to enhance the training of the generative AI model. The analysis results are incorporated as feedback into the generative AI model, adjusting the model's parameters. A specialized training dataset is created based on specific usage patterns and requests.

[0044] Updated model delivery:

[0045] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[0046] Offering user incentives:

[0047] The server rewards users who cooperate with the data analysis. The rewards are applied to the user's account, notified via the user's device, and can be exchanged for benefits such as free use of the next service.

[0048] As a concrete example, when a user uses article generation AI to create a "report on environmental issues," the device records detailed usage information and sends it to the server. The server aggregates and analyzes this data, detecting patterns such as "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings." The server then uses this information to strengthen the learning of the generative AI model and prepares a specialized training dataset. The updated model is sent to the device, and when the user requests article generation again, a higher-quality, more specific report is generated more quickly.

[0049] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0050] The processing flow will be explained below.

[0051] Step 1:

[0052] To use the generative AI, the user launches the application on the device. The device detects the user's actions and begins preparing to record data.

[0053] Step 2:

[0054] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically collects the 5W1H data (when, where, who, what, why, and how) along with the request.

[0055] Step 3:

[0056] The 5W1H data collected by the device is encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[0057] Step 4:

[0058] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[0059] Step 5:

[0060] The server aggregates and categorizes the data stored in the database and applies pattern recognition algorithms to analyze the data, identifying patterns such as frequency of use, regional needs, and usage trends over time.

[0061] Step 6:

[0062] The server uses the analysis results to enhance the training of the generative AI model, incorporates the analysis results as feedback to adjust the model parameters, and creates specialized training datasets based on specific usage patterns and requests.

[0063] Step 7:

[0064] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[0065] Step 8:

[0066] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain highly accurate and specific generative results.

[0067] Step 9:

[0068] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[0069] Through this series of steps, we provide a system that can efficiently utilize the collected data and improve the accuracy and sophistication of generative AI.

[0070] Example 1

[0071] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0072] In current generative AI systems, the process for efficiently collecting and analyzing user usage data and enhancing model learning has not been fully established. As a result, there are limitations on the quality of the generative AI services provided by users, making it difficult to provide highly accurate and sophisticated generated content. Furthermore, if appropriate incentives are not provided for users to provide data, data collection itself may be insufficient. Therefore, a system that solves these issues, efficiently collects and analyzes generative AI usage data, and provides high-quality services is needed.

[0073] 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.

[0074] In this invention, the server includes means for collecting data when a user uses the generative AI, means for encrypting the collected data and transmitting it to the server, means for verifying the transmitted data and securely storing it in a database, means for analyzing the stored data and identifying user behavior patterns and needs, means for enhancing the learning of the generative AI model based on the analysis results, means for delivering the updated generative AI model to the terminal, means for providing rewards to the user, and means for collecting 5W1H data (when, where, who, what, why, and how). This makes it possible to efficiently collect and analyze generative AI usage data and provide users with highly accurate and advanced generative AI services.

[0075] "Means of collecting data" refers to the function by which the terminal automatically detects and records information generated when the user uses the artificial intelligence.

[0076] "Means for encrypting data and sending it to a server" refers to a function that uses encryption technology to securely convert data collected by the terminal and send it safely to a server via the Internet.

[0077] "Means for verifying data and safely storing it in a database" is a function in which the server checks the integrity and security of the data received and then stores it in a database for safekeeping.

[0078] "Means for analyzing stored data and identifying user behavior patterns and needs" refers to a function that analyzes collected and stored data using analytical tools and algorithms to identify user behavior and needs.

[0079] "Means to enhance the learning of the generative AI model based on the analysis results" refers to a function that improves the performance of the model by optimizing and re-learning the parameters of the generative AI model using the results of data analysis.

[0080] "Means for delivering updated generative artificial intelligence models to terminals" refers to a function that sends enhanced AI models from a server to terminals, making the new models available on the user's system.

[0081] The "means for providing rewards to users" is a function for providing incentives such as points or benefits to users who cooperate in providing data.

[0082] "Means of collecting 5W1H data (when, where, who, what, why, and how)" is a function that comprehensively collects detailed information about users' use of AI.

[0083] The present invention relates to a data collection and analysis system that utilizes a generative AI model, and is implemented as follows.

[0084] To implement the invention, a user uses a generation AI service. When the user begins using the service, the device automatically detects the user's actions and begins collecting related data. Specifically, the device collects data such as the start and end times of use, IP address, GPS information, user ID, device ID, type of generation request and specific usage content, purpose information entered by the user, and specific instructions and conditions. This process uses common computing devices such as mobile devices and PCs.

[0085] The collected data is encrypted by the device, for example using AES encryption technology, and sent to a server via HTTPS, a secure communication protocol. The server verifies the integrity and security of the received data and stores it in a database, which may use MongoDB, a NoSQL database.

[0086] The server uses the Python programming language and its data analysis libraries (such as Pandas and Scikit-learn) to analyze the stored data. During the analysis process, the data is aggregated and classified to identify user behavior patterns and needs. For example, a usage analysis might reveal that "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings."

[0087] Based on the analysis results, the server enhances the learning of the generative AI model (e.g., GPT-3). It retrains the model using machine learning libraries such as TensorFlow or PyTorch to optimize parameters. A specialized dataset is created and the model is updated based on this.

[0088] The updated generative AI model is sent from the server to the device via a REST API, which the device receives and applies the new model the next time the user uses the generative AI, allowing the user to receive an improved generative AI service.

[0089] Furthermore, the server has the function of rewarding users who cooperate in providing data. Users are given points and other benefits, which are notified via their terminals. Users can exchange these points for benefits such as discounts on their next usage fee.

[0090] Examples of specific prompts include:

[0091] "Write an academic report of no more than 500 words on an environmental issue. Specific instructions: Cite the latest data and propose a solution to the problem."

[0092] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0093] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0094] Step 1:

[0095] The user begins using the AI ​​generation service.

[0096] Specifically, a user launches the app and clicks the "Create a report on environmental issues" button. The input includes the type of request and its purpose. The output is a request sent to the device.

[0097] Step 2:

[0098] The device detects user actions and collects data.

[0099] Specifically, the terminal collects the start time of use, end time, IP address, GPS information, user ID, device ID, type of generated request, specific usage content, purpose information entered by the user, specific instructions and conditions, etc. The input is the user's actions, and the output is the collected data that is stored locally.

[0100] Step 3:

[0101] The data collected by the device is encrypted and sent to the server.

[0102] Specifically, the data is encrypted using AES encryption technology and sent to the server using the HTTPS protocol. The input is the collected data, and the output is the encrypted data sent to the server.

[0103] Step 4:

[0104] The server validates the received data and stores it securely in a database.

[0105] Specifically, it verifies the data integrity and encryption key, and stores the data in a MongoDB database. The input is the encrypted data, and the output is the verified data stored in the database.

[0106] Step 5:

[0107] The server analyzes the stored data.

[0108] Specifically, we use Python scripts and data analysis libraries (Pandas, Scikit-learn) to aggregate and classify data to identify user behavior patterns and needs. The input is data stored in a database, and the output is analyzed data.

[0109] Step 6:

[0110] The server enhances the learning of the generative AI model based on the analysis results.

[0111] Specifically, TensorFlow or PyTorch are used to retrain generative AI models (e.g., GPT-3) and adjust their parameters. The inputs are analysis results and specialized datasets, and the output is an enhanced generative AI model.

[0112] Step 7:

[0113] The updated generative AI model is distributed from the server to the device.

[0114] Specifically, a new generative AI model is sent to the device via a REST API, and the device receives it and stores it in storage. The input is the updated generative AI model, and the output is the new model stored on the device.

[0115] Step 8:

[0116] The server provides a reward to the user who cooperates in providing data.

[0117] Specifically, points are added to the user's account and notified via the terminal. The input is the data provision record, and the output is the points added to the user's account and notification.

[0118] In this way, through step-by-step processing and specific operations, generative AI services can be implemented efficiently and to a high degree.

[0119] (Application example 1)

[0120] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0121] Currently, the accuracy of customer service and product recommendations in brick-and-mortar stores is insufficient, making it difficult to improve user satisfaction. Furthermore, there is no system in place to effectively collect and analyze customer behavior data and use the results to enhance the model learning of generative artificial intelligence (generative AI). This makes it difficult to achieve highly accurate customer support and product recommendations using generative AI.

[0122] 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.

[0123] In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for enhancing model learning of the generative AI based on the analysis results, means for distributing the updated generative AI model to the terminal, means for collecting data based on customer behavior in a physical store, and means for making optimal product recommendations based on customer behavior information, thereby enabling highly accurate customer support and product recommendations in physical stores.

[0124] "Generative AI" refers to AI that has the ability to automatically generate natural language sentences, images, etc. based on data.

[0125] "Data collection means" refers to devices or software used to collect information about user behavior and events that occur.

[0126] "Means for sending collected data to a server" refers to a mechanism for transferring collected data to a server via the Internet or a dedicated communication protocol.

[0127] "Means for analyzing transmitted data" refers to software that processes the data received by the server, such as classifying, extracting, and recognizing patterns.

[0128] "Means for enhancing model learning" refers to a process for adjusting the parameters of an artificial intelligence model based on the analysis results to improve performance and accuracy.

[0129] The "means for delivering an updated generative AI model to a terminal" is a mechanism for downloading an AI model updated on a server to a user's terminal.

[0130] "Means for collecting data based on customer behavior within a physical store" refers to devices and systems for collecting information such as customer movements, behavior, and purchase history within a physical store.

[0131] "Means for making optimal product recommendations based on customer behavioral information" refers to a system that utilizes collected customer behavioral data to suggest products and services that are suitable for customers.

[0132] This invention provides a generative AI system for providing highly accurate customer service and product recommendations in brick-and-mortar stores. Specifically, this system collects and analyzes customer behavior data and uses the results to strengthen generative AI models.

[0133] Data collection

[0134] Devices (e.g., smart glasses, smartphones) monitor customer behavior in physical stores and collect data, including customer behavior logs (e.g., product searches, purchase history), location information, and usage time.

[0135] Data transmission

[0136] The collected data is then sent from the device to the server, where it is encrypted using the Fernet library and transmitted over a secure communication protocol.

[0137] Data analysis

[0138] The server analyzes the received data, which includes data classification, extraction, and pattern recognition. Analysis identifies customer needs and behavioral patterns, providing insights for optimal product recommendations.

[0139] Enhanced model learning

[0140] The server then uses the analysis results to enhance the generative AI model, adjusting the model's parameters to improve accuracy in the next recommendation or customer support, for example, by adjusting parameters based on trends in a particular time period or region.

[0141] Delivery of updated models

[0142] The updated generative AI model is sent from the server to the device, which then downloads the new model and applies it the next time it is used, ensuring highly accurate support based on the latest information.

[0143] Providing user incentives

[0144] The server will provide rewards to users who cooperate in providing data. These rewards will be applied to the next purchase as points or discounts. Users will also be notified of these benefits via their devices.

[0145] Hardware and software used

[0146] Hardware: Smart glasses, smartphones

[0147] Software: Fernet library (encryption), pattern recognition algorithms, database

[0148] Specific examples

[0149] If a customer uses smart glasses to search for sneakers in a physical store, the prompt would look like this:

[0150] Request: A customer uses smart glasses to search for sneakers.

[0151] Based on these prompts, the system collects and analyzes customer behavior data to enhance and update the generative AI model. The data based on the prompts is sent to a server, and the analysis results are used to adjust the model's parameters. The updated model is then sent to the device and used for the next customer support call.

[0152] This will improve the accuracy of product recommendations and customer support in physical stores, and is expected to increase customer satisfaction.

[0153] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0154] Step 1:

[0155] When a user searches for a product in a physical store using smart glasses or a smartphone, data on the search action is collected. Specifically, the device records the user's search target (product name, category, etc.), location information, timestamp, etc.

[0156] Input: User search behavior, location, timestamp

[0157] Output: Search behavior data (product name, category, location, timestamp)

[0158] Step 2:

[0159] The device encrypts the collected search behavior data using the Fernet library, and then sends the encrypted data to the server.

[0160] Input: Search behavior data

[0161] Output: Encrypted search behavior data

[0162] Step 3:

[0163] The server decrypts the received encrypted data and performs pre-processing for analysis, such as classifying, organizing, and removing noise from the data.

[0164] Input: Encrypted search behavior data

[0165] Output: Preprocessed data

[0166] Step 4:

[0167] The server analyzes the pre-processed data and uses pattern recognition algorithms to identify customer behavior patterns and needs. The results of this analysis are used as feedback for the generative AI model.

[0168] Input: Preprocessed data

[0169] Output: Analysis results (customer behavior patterns, needs, etc.)

[0170] Step 5:

[0171] Based on the analysis results, the server adjusts the parameters of the generative AI model and strengthens model learning, thereby improving the accuracy of the next recommendation.

[0172] Input: Analysis results

[0173] Output: An updated generative AI model

[0174] Step 6:

[0175] The server delivers the updated generative AI model to the device, which then downloads the new model and uses it for the next customer support or product recommendation.

[0176] Input: Updated generative AI model

[0177] Output: A new generative AI model installed on the device.

[0178] Step 7:

[0179] The server provides incentives to users who cooperate in providing data, such as points or rewards, to their accounts and notifies them via their devices.

[0180] Input: User cooperation data

[0181] Output: Reward notification to user (points, bonuses, etc.)

[0182] 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.

[0183] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[0184] Data collection:

[0185] When a user begins using generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed through the emotion engine.

[0186] Data transmission:

[0187] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS). The server verifies the received data and stores it in a database.

[0188] Data Analysis:

[0189] The server analyzes the received data. This analysis includes aggregation, classification, and pattern recognition algorithms for 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[0190] Strengthening model training:

[0191] Based on the data analysis, the server strengthens the learning of the generative AI model. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for each specific emotional state.

[0192] Updated model delivery:

[0193] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[0194] Offering user incentives:

[0195] The server calculates the reward for users who cooperate with the data analysis. The reward is applied to the user's account and notified via the terminal. As a reward, the user receives benefits such as free use of the next service.

[0196] As a concrete example, when a user uses an article generation AI to create a "report on environmental issues," the device analyzes the user's facial expressions and tone of voice using an emotion engine and collects emotional data. The server analyzes this data to understand the user's emotions toward a specific topic. The server then uses this information to strengthen the learning of the generative AI model and generate more appropriate content that reflects the specific emotions.

[0197] In this way, the present invention efficiently collects and analyzes usage data and emotion data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0198] The processing flow will be explained below.

[0199] Step 1:

[0200] To use generative AI, a user launches an application on their device. The device detects the user's actions and begins recording data and collecting emotional data.

[0201] Step 2:

[0202] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically begins collecting the 5W1H data (when, where, who, what, why, and how) along with the request.

[0203] Step 3:

[0204] The 5W1H data collected by the device and the user's emotional data (e.g., facial expression analysis results and tone of voice analysis results) are encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[0205] Step 4:

[0206] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[0207] Step 5:

[0208] The server analyzes the data stored in the database. The analysis includes aggregation, classification, and pattern recognition algorithms of 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage behavior and their emotional state.

[0209] Step 6:

[0210] The server uses the analysis results to enhance the learning of the generative AI model, incorporates the analysis results as feedback to adjust the model's parameters, and creates specialized training datasets tailored to specific usage patterns and emotional states.

[0211] Step 7:

[0212] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[0213] Step 8:

[0214] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain more accurate and specific generative results that also reflect the user's emotional state.

[0215] Step 9:

[0216] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[0217] Through this series of steps, we provide a system that can efficiently utilize the collected emotional data and improve the accuracy and sophistication of the generative AI. As a specific example, when a user uses the article generation AI to create a "report on environmental issues," the device collects the user's facial expressions and tone of voice as emotional data along with usage information and sends it to the server. The server analyzes this data and strengthens the learning of the generative AI model, thereby generating a high-quality report that takes the user's emotions into consideration.

[0218] Example 2

[0219] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0220] Conventional generative AI systems only collect data when users use the AI, and do not consider user emotional data, which limits the accuracy of the generated content. In addition, there is a lack of feedback and incentives for users, making it difficult to effectively strengthen model learning.

[0221] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for strengthening the model learning of the generative AI based on the analysis results, means for delivering the updated generative AI model to the terminal, means for analyzing user emotion data, means for encrypting and transmitting data including the emotion data, and means for collecting 5W1H data. This enables more effective strengthening of the model learning of the generative AI by generating content with high accuracy that reflects the user's emotion data and providing user incentives.

[0222] "User" refers to an individual or organization that utilizes the System to use Generative Artificial Intelligence.

[0223] "Generative AI" refers to an AI model that generates content such as text and images based on given data and prompts.

[0224] "Data collection means" refers to the technology and devices used to collect user usage data, 5W1H data, and emotional data.

[0225] "Server" refers to the computer system that receives, stores, and analyzes collected data to enhance the learning of the generative AI model.

[0226] "Encryption means" refers to the technology or device that encrypts collected data for secure transmission.

[0227] "Emotional data" refers to emotional information analyzed from the user's facial expressions and tone of voice collected through the device's built-in camera and microphone, or external devices.

[0228] "Emotion engine" refers to software or hardware that analyzes collected emotion data and recognizes the user's emotional state.

[0229] "5W1H data" refers to data that includes information on when, where, who, what, why, and how.

[0230] "Reward means" refers to the technology or mechanism that provides rewards to users who cooperate with data analysis.

[0231] A "generative AI model" refers to an artificial intelligence model that generates content such as text and images based on user input data and analysis results.

[0232] "Model learning enhancement means" refers to the techniques and processes that improve and update the generative AI model based on received data and analysis results.

[0233] "Terminal" refers to a device or equipment used by a user, and may include sensors such as a camera or microphone.

[0234] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[0235] First, when a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). This emotional data is acquired using the device's built-in camera and microphone, or an external device. The collected emotional data is analyzed by an emotion engine running on the device. This analysis identifies the user's emotional state.

[0236] The collected 5W1H data and emotion data are encrypted by the device and sent to a server using a secure communication protocol (e.g., HTTPS), where the server verifies the received data and stores it in a database.

[0237] The server then analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state. Specifically, detailed analysis results can be obtained, such as frequency of use, needs by region, and usage trends by time of day.

[0238] The server strengthens the learning of the generative AI model based on these analysis results. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's response and generated content. For example, when a user feels "concerned," the generative AI can adjust its content to provide more reassuring content. This creates a specialized learning dataset tailored to a specific emotional state.

[0239] The updated generative AI model is sent from the server to the device, which then downloads the updated AI model and applies it the next time the user uses the generative AI, providing a more accurate service.

[0240] In addition, for users who cooperate with data analysis, the server calculates a reward and applies it to the user's account. This reward is notified to the user via their device, and the user can receive benefits such as free use of the service next time.

[0241] A specific example is when a user uses an article generation AI to create a "report on environmental issues." In this case, when the user begins working, the device detects their usage and collects information such as the start time and location. At the same time, the emotion engine analyzes the user's facial expressions and tone of voice via the camera and microphone, collecting emotional data. The server analyzes this data to understand how the user feels about a particular topic. Based on the results of this analysis, the generative AI model strengthens its learning so that it can generate appropriate content that reflects specific emotions.

[0242] Here is an example prompt:

[0243] "We're using generative AI to create a report on environmental issues. We've analyzed the user sentiments and found that 'interesting' and 'concerned' are the two most prominent. Based on this, we'd like you to generate a bullet point summary that reflects those sentiments."

[0244] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0245] Step 1:

[0246] The user begins using the generation AI. The device detects the user's actions. The input is the user's action (for example, launching the article generation AI), and the output is data such as the time and location of use. Specifically, the device records the user's action log.

[0247] Step 2:

[0248] The device collects 5W1H data and emotional data. The input is the user's usage situation (when, where, who, what, why, how) as well as the user's facial expression and tone of voice, and the output is analyzed emotional data and 5W1H data. Specifically, the device uses the built-in camera and microphone to collect data on the user's facial expression and voice, which is then analyzed by the emotion engine.

[0249] Step 3:

[0250] The device encrypts the data collected and analyzed. The input is 5W1H data and analyzed emotional data, and the output is encrypted data. Specifically, the encryption module in the device encrypts the data and converts it into a secure format.

[0251] Step 4:

[0252] The terminal sends encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is the encrypted data, and the output is the completion of data transmission to the server. Specifically, the terminal's communication module sends the data to the server.

[0253] Step 5:

[0254] The server verifies the data it receives and stores it in a database. The input is the encrypted data sent from the device, and the output is data in an analyzable format stored in the database. The specific operation is that the server checks the integrity of the data and stores it in the database.

[0255] Step 6:

[0256] The server analyzes the stored data. The input is the 5W1H data and emotional data stored in the database, and the output is an analysis result that clarifies the relationship between the user's usage and emotional state. Specifically, the server runs aggregation, classification, and pattern recognition algorithms to analyze the data.

[0257] Step 7:

[0258] The server strengthens the learning of the generative AI model based on the analysis results. The input is the analyzed data on the user's usage and emotional state, and the output is an improved generative AI model. Specifically, the server adjusts the parameters of the generative AI model and updates the model's learning.

[0259] Step 8:

[0260] The server delivers the updated generative AI model to the device. The input is the updated generative AI model, and the output is the completion of model delivery to the device. Specifically, the server sends the new generative AI model to the device, which receives and stores it.

[0261] Step 9:

[0262] The server calculates rewards for users who cooperate with data analysis and notifies the users via their devices. The input is the user's cooperation data, and the output is reward information applied to the user's account. Specifically, the server calculates the reward and notifies the result to the user's device.

[0263] Step 10:

[0264] The device downloads the new generative AI model and applies it the next time the user uses the generative AI. The input is the delivered generative AI model, and the output is content generation using the updated generative AI. Specifically, the device saves the new model and applies it the next time the user uses the generative AI.

[0265] Through the above processing steps, we have created a system that uses user emotion data to improve the accuracy of generative AI models.

[0266] (Application example 2)

[0267] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0268] The challenge is to provide more accurate services by efficiently utilizing user emotional data in model training for generative AI. It is also desirable to improve user satisfaction and increase the utilization rate of services by dynamically providing optimal information according to the user's emotional state.

[0269] The specific processing by the specific processing unit 290 of the data processing device 12 in application example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses a generative AI, means for transmitting the collected data and the user's emotional data to the server, means for analyzing the transmitted data and emotional data, means for strengthening model learning of the generative AI based on the analysis results, means for delivering an updated generative AI model to the terminal, and means for selecting and displaying information based on the user's emotional data. This enables a highly accurate generative AI service that reflects the user's emotions.

[0270] "Means for collecting data when a user uses a generative artificial intelligence" refers to a device or software that has the function of recording the actions and inputs a user makes when using a generative artificial intelligence system.

[0271] The "means for transmitting collected data and user emotional data to a server" refers to a device or software that has the function of sending the acquired user usage data and emotional state data to a server using a secure communication protocol.

[0272] "Means for analyzing transmitted data and emotional data" refers to algorithms and software that classify and aggregate the data received by the server and perform detailed analysis, including of the user's emotional state.

[0273] The "means for strengthening the model learning of generative AI based on the analysis results" is a system that has the function of adjusting the parameters of the generative AI model and increasing its adaptability by using user data and emotional data obtained through the analysis.

[0274] "Means for delivering updated generative AI models to terminals" refers to a system that has the function of sending a generative AI model that reflects the latest learning content to a terminal, allowing the user to use the updated model the next time they use it.

[0275] "Means for selecting and displaying information based on user emotional data" refers to a system that analyzes a user's emotional state data, selects the most appropriate information or advertisements, and displays them on the user's device.

[0276] "5W1H data" is a data format that refers to information on when, where, who, what, why, and how.

[0277] This invention is a system that combines data obtained when a user uses generative artificial intelligence (generative AI) with an emotion engine that recognizes the user's emotions to strengthen the model learning of the generative AI and provide a more accurate service. Below, we will explain how the server, terminal, and user each implement this system.

[0278] Data collection:

[0279] When a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed by the emotion engine. This allows for detailed records of the user's usage and their emotional state at the time.

[0280] Data transmission:

[0281] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS), where it is verified and stored in a database.

[0282] Data Analysis:

[0283] The server analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[0284] Strengthening model training:

[0285] The server strengthens the learning of the generative AI model based on data analysis. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for a specific emotional state.

[0286] Updated model delivery:

[0287] The server delivers the generative AI model that reflects the latest learning content to the device, and the device downloads the new AI model and applies it the next time the user uses the generative AI.

[0288] Emotional Adaptation Display:

[0289] Based on the user's emotional data, the system selects the most appropriate information and advertisements and displays them on devices such as smart glasses, thereby providing the most effective advertisements and information to users in real time.

[0290] For example, if a user is wearing smart glasses and the camera detects the user smiling, the application will interpret this as "happiness." Based on this emotional information, the server will select the most appropriate advertisement, such as an advertisement for "ice cream on sale," and display it on the smart glasses' display. In this way, the optimal advertisement is dynamically provided according to the user's emotional state.

[0291] Example prompt sentence:

[0292] Generate the perfect ad to show when the user is smiling.

[0293] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0294] Step 1:

[0295] Data collection:

[0296] When a user begins using the generative AI, the device automatically starts up and detects the user's actions. Specifically, the device's built-in camera and microphone are used to collect the user's 5W1H data as well as emotional data. The input data is the user's voice, facial expressions, and behavior, which are analyzed by the emotion engine to output the user's emotional state.

[0297] Step 2:

[0298] Data transmission:

[0299] The collected 5W1H data and emotion data are encrypted by the device. The device then sends this encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is a set of emotion and 5W1H data, and the output is an encrypted data stream.

[0300] Step 3:

[0301] Data reception and validation:

[0302] The server decrypts and verifies the data it receives, which involves the process of ensuring that the data entered is accurate and complete. The input is the encrypted data set, and the output is the decrypted data.

[0303] Step 4:

[0304] Data Retention:

[0305] The server then stores the validated data in a database. This process preserves the integrity of the data and properly categorizes and stores it for future analysis and learning. The input is the decrypted data, and the output is a stored entry in the database.

[0306] Step 5:

[0307] Data Analysis:

[0308] The server analyzes the data stored in the database. This analysis process involves aggregating, classifying, and recognizing patterns from the 5W1H data, and combining it with emotional data to find statistical relationships. The input is the raw data retrieved from the database, and the output is an analysis report or pattern recognition results.

[0309] Step 6:

[0310] Strengthening model training:

[0311] Based on the analysis results, the server enhances the learning of the generative AI model, especially the process of optimizing the model parameters by reflecting the user's emotional data. The input is the analysis report and the existing model parameters, and the output is the optimized generative AI model.

[0312] Step 7:

[0313] Updated model delivery:

[0314] The server delivers a generative AI model incorporating the latest learning results to the device. The device downloads this model and applies the new model the next time the user uses the generative AI. The input is the optimized generative AI model, and the output is the new AI model downloaded to the device.

[0315] Step 8:

[0316] Emotional Adaptation Display:

[0317] When a user uses smart glasses, for example, the system selects the most appropriate information or advertisement based on emotional data and displays it in real time. In this process, an algorithm operates to select specific advertisements or information based on the user's current emotional state. The input is emotional data obtained in real time, and the output is advertisements or information displayed on the user interface.

[0318] For example, when the user smiles, an advertisement is selected based on the prompt, "Generate the best advertisement to display when the user is smiling," and an "ice cream sale advertisement" is displayed on the smart glasses display.

[0319] 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.

[0320] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0321] 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.

[0322] [Second embodiment]

[0323] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0324] 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.

[0325] 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).

[0326] 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.

[0327] 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.

[0328] 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).

[0329] 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.

[0330] 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.

[0331] 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.

[0332] 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.

[0333] 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.

[0334] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0335] This invention relates to a system that provides more accurate generative AI services by collecting data on the use of generative AI and enhancing the learning of AI models through analysis. Below, we will explain how this system is implemented by the server, terminal, and user.

[0336] Data collection:

[0337] When a user begins using the generation AI, the device automatically detects the user's actions and begins recording data. The device collects data such as the start and end times of use (when), IP address and GPS information (where), user ID and device ID (who), the type of generation request and specific usage (what), purpose information entered by the user (why), and specific instructions and conditions (how).

[0338] Data transmission:

[0339] The collected data is encrypted by the device and sent using a secure communication protocol to a server, which verifies the received data and stores it in a database.

[0340] Data Analysis:

[0341] The server analyzes the received data using aggregation, classification, and pattern recognition algorithms to identify usage frequency, regional needs, and usage trends by time of day, resulting in a detailed analysis of the use of the Generative AI.

[0342] Strengthening model training:

[0343] The server uses data analysis to enhance the training of the generative AI model. The analysis results are incorporated as feedback into the generative AI model, adjusting the model's parameters. A specialized training dataset is created based on specific usage patterns and requests.

[0344] Updated model delivery:

[0345] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[0346] Offering user incentives:

[0347] The server rewards users who cooperate with the data analysis. The rewards are applied to the user's account, notified via the user's device, and can be exchanged for benefits such as free use of the next service.

[0348] As a concrete example, when a user uses article generation AI to create a "report on environmental issues," the device records detailed usage information and sends it to the server. The server aggregates and analyzes this data, detecting patterns such as "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings." The server then uses this information to strengthen the learning of the generative AI model and prepares a specialized training dataset. The updated model is sent to the device, and when the user requests article generation again, a higher-quality, more specific report is generated more quickly.

[0349] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0350] The processing flow will be explained below.

[0351] Step 1:

[0352] To use the generative AI, the user launches the application on the device. The device detects the user's actions and begins preparing to record data.

[0353] Step 2:

[0354] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically collects the 5W1H data (when, where, who, what, why, and how) along with the request.

[0355] Step 3:

[0356] The 5W1H data collected by the device is encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[0357] Step 4:

[0358] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[0359] Step 5:

[0360] The server aggregates and categorizes the data stored in the database and applies pattern recognition algorithms to analyze the data, identifying patterns such as frequency of use, regional needs, and usage trends over time.

[0361] Step 6:

[0362] The server uses the analysis results to enhance the training of the generative AI model, incorporates the analysis results as feedback to adjust the model parameters, and creates specialized training datasets based on specific usage patterns and requests.

[0363] Step 7:

[0364] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[0365] Step 8:

[0366] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain highly accurate and specific generative results.

[0367] Step 9:

[0368] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[0369] Through this series of steps, we provide a system that can efficiently utilize the collected data and improve the accuracy and sophistication of generative AI.

[0370] Example 1

[0371] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0372] In current generative AI systems, the process for efficiently collecting and analyzing user usage data and enhancing model learning has not been fully established. As a result, there are limitations on the quality of the generative AI services provided by users, making it difficult to provide highly accurate and sophisticated generated content. Furthermore, if appropriate incentives are not provided for users to provide data, data collection itself may be insufficient. Therefore, a system that solves these issues, efficiently collects and analyzes generative AI usage data, and provides high-quality services is needed.

[0373] 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.

[0374] In this invention, the server includes means for collecting data when a user uses the generative AI, means for encrypting the collected data and transmitting it to the server, means for verifying the transmitted data and securely storing it in a database, means for analyzing the stored data and identifying user behavior patterns and needs, means for enhancing the learning of the generative AI model based on the analysis results, means for delivering the updated generative AI model to the terminal, means for providing rewards to the user, and means for collecting 5W1H data (when, where, who, what, why, and how). This makes it possible to efficiently collect and analyze generative AI usage data and provide users with highly accurate and advanced generative AI services.

[0375] "Means of collecting data" refers to the function by which the terminal automatically detects and records information generated when the user uses the artificial intelligence.

[0376] "Means for encrypting data and sending it to a server" refers to a function that uses encryption technology to securely convert data collected by the terminal and send it safely to a server via the Internet.

[0377] "Means for verifying data and safely storing it in a database" is a function in which the server checks the integrity and security of the data received and then stores it in a database for safekeeping.

[0378] "Means for analyzing stored data and identifying user behavior patterns and needs" refers to a function that analyzes collected and stored data using analytical tools and algorithms to identify user behavior and needs.

[0379] "Means to enhance the learning of the generative AI model based on the analysis results" refers to a function that improves the performance of the model by optimizing and re-learning the parameters of the generative AI model using the results of data analysis.

[0380] "Means for delivering updated generative artificial intelligence models to terminals" refers to a function that sends enhanced AI models from a server to terminals, making the new models available on the user's system.

[0381] The "means for providing rewards to users" is a function for providing incentives such as points or benefits to users who cooperate in providing data.

[0382] "Means of collecting 5W1H data (when, where, who, what, why, and how)" is a function that comprehensively collects detailed information about users' use of AI.

[0383] The present invention relates to a data collection and analysis system that utilizes a generative AI model, and is implemented as follows.

[0384] To implement the invention, a user uses a generation AI service. When the user begins using the service, the device automatically detects the user's actions and begins collecting related data. Specifically, the device collects data such as the start and end times of use, IP address, GPS information, user ID, device ID, type of generation request and specific usage content, purpose information entered by the user, and specific instructions and conditions. This process uses common computing devices such as mobile devices and PCs.

[0385] The collected data is encrypted by the device, for example using AES encryption technology, and sent to a server via HTTPS, a secure communication protocol. The server verifies the integrity and security of the received data and stores it in a database, which may use MongoDB, a NoSQL database.

[0386] The server uses the Python programming language and its data analysis libraries (such as Pandas and Scikit-learn) to analyze the stored data. During the analysis process, the data is aggregated and classified to identify user behavior patterns and needs. For example, a usage analysis might reveal that "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings."

[0387] Based on the analysis results, the server enhances the learning of the generative AI model (e.g., GPT-3). It retrains the model using machine learning libraries such as TensorFlow or PyTorch to optimize parameters. A specialized dataset is created and the model is updated based on this.

[0388] The updated generative AI model is sent from the server to the device via a REST API, which the device receives and applies the new model the next time the user uses the generative AI, allowing the user to receive an improved generative AI service.

[0389] Furthermore, the server has the function of rewarding users who cooperate in providing data. Users are given points and other benefits, which are notified via their terminals. Users can exchange these points for benefits such as discounts on their next usage fee.

[0390] Examples of specific prompts include:

[0391] "Write an academic report of no more than 500 words on an environmental issue. Specific instructions: Cite the latest data and propose a solution to the problem."

[0392] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0393] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0394] Step 1:

[0395] The user begins using the AI ​​generation service.

[0396] Specifically, a user launches the app and clicks the "Create a report on environmental issues" button. The input includes the type of request and its purpose. The output is a request sent to the device.

[0397] Step 2:

[0398] The device detects user actions and collects data.

[0399] Specifically, the terminal collects the start time of use, end time, IP address, GPS information, user ID, device ID, type of generated request, specific usage content, purpose information entered by the user, specific instructions and conditions, etc. The input is the user's actions, and the output is the collected data that is stored locally.

[0400] Step 3:

[0401] The data collected by the device is encrypted and sent to the server.

[0402] Specifically, the data is encrypted using AES encryption technology and sent to the server using the HTTPS protocol. The input is the collected data, and the output is the encrypted data sent to the server.

[0403] Step 4:

[0404] The server validates the received data and stores it securely in a database.

[0405] Specifically, it verifies the data integrity and encryption key, and stores the data in a MongoDB database. The input is the encrypted data, and the output is the verified data stored in the database.

[0406] Step 5:

[0407] The server analyzes the stored data.

[0408] Specifically, we use Python scripts and data analysis libraries (Pandas, Scikit-learn) to aggregate and classify data to identify user behavior patterns and needs. The input is data stored in a database, and the output is analyzed data.

[0409] Step 6:

[0410] The server enhances the learning of the generative AI model based on the analysis results.

[0411] Specifically, TensorFlow or PyTorch are used to retrain generative AI models (e.g., GPT-3) and adjust their parameters. The inputs are analysis results and specialized datasets, and the output is an enhanced generative AI model.

[0412] Step 7:

[0413] The updated generative AI model is distributed from the server to the device.

[0414] Specifically, a new generative AI model is sent to the device via a REST API, and the device receives it and stores it in storage. The input is the updated generative AI model, and the output is the new model stored on the device.

[0415] Step 8:

[0416] The server provides a reward to the user who cooperates in providing data.

[0417] Specifically, points are added to the user's account and notified via the terminal. The input is the data provision record, and the output is the points added to the user's account and notification.

[0418] In this way, through step-by-step processing and specific operations, generative AI services can be implemented efficiently and to a high degree.

[0419] (Application example 1)

[0420] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0421] Currently, the accuracy of customer service and product recommendations in brick-and-mortar stores is insufficient, making it difficult to improve user satisfaction. Furthermore, there is no system in place to effectively collect and analyze customer behavior data and use the results to enhance the model learning of generative artificial intelligence (generative AI). This makes it difficult to achieve highly accurate customer support and product recommendations using generative AI.

[0422] 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.

[0423] In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for enhancing model learning of the generative AI based on the analysis results, means for distributing the updated generative AI model to the terminal, means for collecting data based on customer behavior in a physical store, and means for making optimal product recommendations based on customer behavior information, thereby enabling highly accurate customer support and product recommendations in physical stores.

[0424] "Generative AI" refers to AI that has the ability to automatically generate natural language sentences, images, etc. based on data.

[0425] "Data collection means" refers to devices or software used to collect information about user behavior and events that occur.

[0426] "Means for sending collected data to a server" refers to a mechanism for transferring collected data to a server via the Internet or a dedicated communication protocol.

[0427] "Means for analyzing transmitted data" refers to software that processes the data received by the server, such as classifying, extracting, and recognizing patterns.

[0428] "Means for enhancing model learning" refers to a process for adjusting the parameters of an artificial intelligence model based on the analysis results to improve performance and accuracy.

[0429] The "means for delivering an updated generative AI model to a terminal" is a mechanism for downloading an AI model updated on a server to a user's terminal.

[0430] "Means for collecting data based on customer behavior within a physical store" refers to devices and systems for collecting information such as customer movements, behavior, and purchase history within a physical store.

[0431] "Means for making optimal product recommendations based on customer behavioral information" refers to a system that utilizes collected customer behavioral data to suggest products and services that are suitable for customers.

[0432] This invention provides a generative AI system for providing highly accurate customer service and product recommendations in brick-and-mortar stores. Specifically, this system collects and analyzes customer behavior data and uses the results to strengthen generative AI models.

[0433] Data collection

[0434] Devices (e.g., smart glasses, smartphones) monitor customer behavior in physical stores and collect data, including customer behavior logs (e.g., product searches, purchase history), location information, and usage time.

[0435] Data transmission

[0436] The collected data is then sent from the device to the server, where it is encrypted using the Fernet library and transmitted over a secure communication protocol.

[0437] Data analysis

[0438] The server analyzes the received data, which includes data classification, extraction, and pattern recognition. Analysis identifies customer needs and behavioral patterns, providing insights for optimal product recommendations.

[0439] Enhanced model learning

[0440] The server then uses the analysis results to enhance the generative AI model, adjusting the model's parameters to improve accuracy in the next recommendation or customer support, for example, by adjusting parameters based on trends in a particular time period or region.

[0441] Delivery of updated models

[0442] The updated generative AI model is sent from the server to the device, which then downloads the new model and applies it the next time it is used, ensuring highly accurate support based on the latest information.

[0443] Providing user incentives

[0444] The server will provide rewards to users who cooperate in providing data. These rewards will be applied to the next purchase as points or discounts. Users will also be notified of these benefits via their devices.

[0445] Hardware and software used

[0446] Hardware: Smart glasses, smartphones

[0447] Software: Fernet library (encryption), pattern recognition algorithms, database

[0448] Specific examples

[0449] If a customer uses smart glasses to search for sneakers in a physical store, the prompt would look like this:

[0450] Request: A customer uses smart glasses to search for sneakers.

[0451] Based on these prompts, the system collects and analyzes customer behavior data to enhance and update the generative AI model. The data based on the prompts is sent to a server, and the analysis results are used to adjust the model's parameters. The updated model is then sent to the device and used for the next customer support call.

[0452] This will improve the accuracy of product recommendations and customer support in physical stores, and is expected to increase customer satisfaction.

[0453] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0454] Step 1:

[0455] When a user searches for a product in a physical store using smart glasses or a smartphone, data on the search action is collected. Specifically, the device records the user's search target (product name, category, etc.), location information, timestamp, etc.

[0456] Input: User search behavior, location, timestamp

[0457] Output: Search behavior data (product name, category, location, timestamp)

[0458] Step 2:

[0459] The device encrypts the collected search behavior data using the Fernet library, and then sends the encrypted data to the server.

[0460] Input: Search behavior data

[0461] Output: Encrypted search behavior data

[0462] Step 3:

[0463] The server decrypts the received encrypted data and performs pre-processing for analysis, such as classifying, organizing, and removing noise from the data.

[0464] Input: Encrypted search behavior data

[0465] Output: Preprocessed data

[0466] Step 4:

[0467] The server analyzes the pre-processed data and uses pattern recognition algorithms to identify customer behavior patterns and needs. The results of this analysis are used as feedback for the generative AI model.

[0468] Input: Preprocessed data

[0469] Output: Analysis results (customer behavior patterns, needs, etc.)

[0470] Step 5:

[0471] Based on the analysis results, the server adjusts the parameters of the generative AI model and strengthens model learning, thereby improving the accuracy of the next recommendation.

[0472] Input: Analysis results

[0473] Output: An updated generative AI model

[0474] Step 6:

[0475] The server delivers the updated generative AI model to the device, which then downloads the new model and uses it for the next customer support or product recommendation.

[0476] Input: Updated generative AI model

[0477] Output: A new generative AI model installed on the device.

[0478] Step 7:

[0479] The server provides incentives to users who cooperate in providing data, such as points or rewards, to their accounts and notifies them via their devices.

[0480] Input: User cooperation data

[0481] Output: Reward notification to user (points, bonuses, etc.)

[0482] 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.

[0483] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[0484] Data collection:

[0485] When a user begins using generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed through the emotion engine.

[0486] Data transmission:

[0487] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS). The server verifies the received data and stores it in a database.

[0488] Data Analysis:

[0489] The server analyzes the received data. This analysis includes aggregation, classification, and pattern recognition algorithms for 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[0490] Strengthening model training:

[0491] Based on the data analysis, the server strengthens the learning of the generative AI model. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for each specific emotional state.

[0492] Updated model delivery:

[0493] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[0494] Offering user incentives:

[0495] The server calculates the reward for users who cooperate with the data analysis. The reward is applied to the user's account and notified via the terminal. As a reward, the user receives benefits such as free use of the next service.

[0496] As a concrete example, when a user uses an article generation AI to create a "report on environmental issues," the device analyzes the user's facial expressions and tone of voice using an emotion engine and collects emotional data. The server analyzes this data to understand the user's emotions toward a specific topic. The server then uses this information to strengthen the learning of the generative AI model and generate more appropriate content that reflects the specific emotions.

[0497] In this way, the present invention efficiently collects and analyzes usage data and emotion data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0498] The processing flow will be explained below.

[0499] Step 1:

[0500] To use generative AI, a user launches an application on their device. The device detects the user's actions and begins recording data and collecting emotional data.

[0501] Step 2:

[0502] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically begins collecting the 5W1H data (when, where, who, what, why, and how) along with the request.

[0503] Step 3:

[0504] The 5W1H data collected by the device and the user's emotional data (e.g., facial expression analysis results and tone of voice analysis results) are encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[0505] Step 4:

[0506] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[0507] Step 5:

[0508] The server analyzes the data stored in the database. The analysis includes aggregation, classification, and pattern recognition algorithms of 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage behavior and their emotional state.

[0509] Step 6:

[0510] The server uses the analysis results to enhance the learning of the generative AI model, incorporates the analysis results as feedback to adjust the model's parameters, and creates specialized training datasets tailored to specific usage patterns and emotional states.

[0511] Step 7:

[0512] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[0513] Step 8:

[0514] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain more accurate and specific generative results that also reflect the user's emotional state.

[0515] Step 9:

[0516] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[0517] Through this series of steps, we provide a system that can efficiently utilize the collected emotional data and improve the accuracy and sophistication of the generative AI. As a specific example, when a user uses the article generation AI to create a "report on environmental issues," the device collects the user's facial expressions and tone of voice as emotional data along with usage information and sends it to the server. The server analyzes this data and strengthens the learning of the generative AI model, thereby generating a high-quality report that takes the user's emotions into consideration.

[0518] Example 2

[0519] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0520] Conventional generative AI systems only collect data when users use the AI, and do not consider user emotional data, which limits the accuracy of the generated content. In addition, there is a lack of feedback and incentives for users, making it difficult to effectively strengthen model learning.

[0521] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for strengthening the model learning of the generative AI based on the analysis results, means for delivering the updated generative AI model to the terminal, means for analyzing user emotion data, means for encrypting and transmitting data including the emotion data, and means for collecting 5W1H data. This enables more effective strengthening of the model learning of the generative AI by generating content with high accuracy that reflects the user's emotion data and providing user incentives.

[0522] "User" refers to an individual or organization that utilizes the System to use Generative Artificial Intelligence.

[0523] "Generative AI" refers to an AI model that generates content such as text and images based on given data and prompts.

[0524] "Data collection means" refers to the technology and devices used to collect user usage data, 5W1H data, and emotional data.

[0525] "Server" refers to the computer system that receives, stores, and analyzes collected data to enhance the learning of the generative AI model.

[0526] "Encryption means" refers to the technology or device that encrypts collected data for secure transmission.

[0527] "Emotional data" refers to emotional information analyzed from the user's facial expressions and tone of voice collected through the device's built-in camera and microphone, or external devices.

[0528] "Emotion engine" refers to software or hardware that analyzes collected emotion data and recognizes the user's emotional state.

[0529] "5W1H data" refers to data that includes information on when, where, who, what, why, and how.

[0530] "Reward means" refers to the technology or mechanism that provides rewards to users who cooperate with data analysis.

[0531] A "generative AI model" refers to an artificial intelligence model that generates content such as text and images based on user input data and analysis results.

[0532] "Model learning enhancement means" refers to the techniques and processes that improve and update the generative AI model based on received data and analysis results.

[0533] "Terminal" refers to a device or equipment used by a user, and may include sensors such as a camera or microphone.

[0534] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[0535] First, when a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). This emotional data is acquired using the device's built-in camera and microphone, or an external device. The collected emotional data is analyzed by an emotion engine running on the device. This analysis identifies the user's emotional state.

[0536] The collected 5W1H data and emotion data are encrypted by the device and sent to a server using a secure communication protocol (e.g., HTTPS), where the server verifies the received data and stores it in a database.

[0537] The server then analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state. Specifically, detailed analysis results can be obtained, such as frequency of use, needs by region, and usage trends by time of day.

[0538] The server strengthens the learning of the generative AI model based on these analysis results. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's response and generated content. For example, when a user feels "concerned," the generative AI can adjust its content to provide more reassuring content. This creates a specialized learning dataset tailored to a specific emotional state.

[0539] The updated generative AI model is sent from the server to the device, which then downloads the updated AI model and applies it the next time the user uses the generative AI, providing a more accurate service.

[0540] In addition, for users who cooperate with data analysis, the server calculates a reward and applies it to the user's account. This reward is notified to the user via their device, and the user can receive benefits such as free use of the service next time.

[0541] A specific example is when a user uses an article generation AI to create a "report on environmental issues." In this case, when the user begins working, the device detects their usage and collects information such as the start time and location. At the same time, the emotion engine analyzes the user's facial expressions and tone of voice via the camera and microphone, collecting emotional data. The server analyzes this data to understand how the user feels about a particular topic. Based on the results of this analysis, the generative AI model strengthens its learning so that it can generate appropriate content that reflects specific emotions.

[0542] Here is an example prompt:

[0543] "We're using generative AI to create a report on environmental issues. We've analyzed the user sentiments and found that 'interesting' and 'concerned' are the two most prominent. Based on this, we'd like you to generate a bullet point summary that reflects those sentiments."

[0544] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0545] Step 1:

[0546] The user begins using the generation AI. The device detects the user's actions. The input is the user's action (for example, launching the article generation AI), and the output is data such as the time and location of use. Specifically, the device records the user's action log.

[0547] Step 2:

[0548] The device collects 5W1H data and emotional data. The input is the user's usage situation (when, where, who, what, why, how) as well as the user's facial expression and tone of voice, and the output is analyzed emotional data and 5W1H data. Specifically, the device uses the built-in camera and microphone to collect data on the user's facial expression and voice, which is then analyzed by the emotion engine.

[0549] Step 3:

[0550] The device encrypts the data collected and analyzed. The input is 5W1H data and analyzed emotional data, and the output is encrypted data. Specifically, the encryption module in the device encrypts the data and converts it into a secure format.

[0551] Step 4:

[0552] The terminal sends encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is the encrypted data, and the output is the completion of data transmission to the server. Specifically, the terminal's communication module sends the data to the server.

[0553] Step 5:

[0554] The server verifies the data it receives and stores it in a database. The input is the encrypted data sent from the device, and the output is data in an analyzable format stored in the database. The specific operation is that the server checks the integrity of the data and stores it in the database.

[0555] Step 6:

[0556] The server analyzes the stored data. The input is the 5W1H data and emotional data stored in the database, and the output is an analysis result that clarifies the relationship between the user's usage and emotional state. Specifically, the server runs aggregation, classification, and pattern recognition algorithms to analyze the data.

[0557] Step 7:

[0558] The server strengthens the learning of the generative AI model based on the analysis results. The input is the analyzed data on the user's usage and emotional state, and the output is an improved generative AI model. Specifically, the server adjusts the parameters of the generative AI model and updates the model's learning.

[0559] Step 8:

[0560] The server delivers the updated generative AI model to the device. The input is the updated generative AI model, and the output is the completion of model delivery to the device. Specifically, the server sends the new generative AI model to the device, which receives and stores it.

[0561] Step 9:

[0562] The server calculates rewards for users who cooperate with data analysis and notifies the users via their devices. The input is the user's cooperation data, and the output is reward information applied to the user's account. Specifically, the server calculates the reward and notifies the result to the user's device.

[0563] Step 10:

[0564] The device downloads the new generative AI model and applies it the next time the user uses the generative AI. The input is the delivered generative AI model, and the output is content generation using the updated generative AI. Specifically, the device saves the new model and applies it the next time the user uses the generative AI.

[0565] Through the above processing steps, we have created a system that uses user emotion data to improve the accuracy of generative AI models.

[0566] (Application example 2)

[0567] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0568] The challenge is to provide more accurate services by efficiently utilizing user emotional data in model training for generative AI. It is also desirable to improve user satisfaction and increase the utilization rate of services by dynamically providing optimal information according to the user's emotional state.

[0569] The specific processing by the specific processing unit 290 of the data processing device 12 in application example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses a generative AI, means for transmitting the collected data and the user's emotional data to the server, means for analyzing the transmitted data and emotional data, means for strengthening model learning of the generative AI based on the analysis results, means for delivering an updated generative AI model to the terminal, and means for selecting and displaying information based on the user's emotional data. This enables a highly accurate generative AI service that reflects the user's emotions.

[0570] "Means for collecting data when a user uses a generative artificial intelligence" refers to a device or software that has the function of recording the actions and inputs a user makes when using a generative artificial intelligence system.

[0571] The "means for transmitting collected data and user emotional data to a server" refers to a device or software that has the function of sending the acquired user usage data and emotional state data to a server using a secure communication protocol.

[0572] "Means for analyzing transmitted data and emotional data" refers to algorithms and software that classify and aggregate the data received by the server and perform detailed analysis, including of the user's emotional state.

[0573] The "means for strengthening the model learning of generative AI based on the analysis results" is a system that has the function of adjusting the parameters of the generative AI model and increasing its adaptability by using user data and emotional data obtained through the analysis.

[0574] "Means for delivering updated generative AI models to terminals" refers to a system that has the function of sending a generative AI model that reflects the latest learning content to a terminal, allowing the user to use the updated model the next time they use it.

[0575] "Means for selecting and displaying information based on user emotional data" refers to a system that analyzes a user's emotional state data, selects the most appropriate information or advertisements, and displays them on the user's device.

[0576] "5W1H data" is a data format that refers to information on when, where, who, what, why, and how.

[0577] This invention is a system that combines data obtained when a user uses generative artificial intelligence (generative AI) with an emotion engine that recognizes the user's emotions to strengthen the model learning of the generative AI and provide a more accurate service. Below, we will explain how the server, terminal, and user each implement this system.

[0578] Data collection:

[0579] When a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed by the emotion engine. This allows for detailed records of the user's usage and their emotional state at the time.

[0580] Data transmission:

[0581] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS), where it is verified and stored in a database.

[0582] Data Analysis:

[0583] The server analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[0584] Strengthening model training:

[0585] The server strengthens the learning of the generative AI model based on data analysis. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for a specific emotional state.

[0586] Updated model delivery:

[0587] The server delivers the generative AI model that reflects the latest learning content to the device, and the device downloads the new AI model and applies it the next time the user uses the generative AI.

[0588] Emotional Adaptation Display:

[0589] Based on the user's emotional data, the system selects the most appropriate information and advertisements and displays them on devices such as smart glasses, thereby providing the most effective advertisements and information to users in real time.

[0590] For example, if a user is wearing smart glasses and the camera detects the user smiling, the application will interpret this as "happiness." Based on this emotional information, the server will select the most appropriate advertisement, such as an advertisement for "ice cream on sale," and display it on the smart glasses' display. In this way, the optimal advertisement is dynamically provided according to the user's emotional state.

[0591] Example prompt sentence:

[0592] Generate the perfect ad to show when the user is smiling.

[0593] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0594] Step 1:

[0595] Data collection:

[0596] When a user begins using the generative AI, the device automatically starts up and detects the user's actions. Specifically, the device's built-in camera and microphone are used to collect the user's 5W1H data as well as emotional data. The input data is the user's voice, facial expressions, and behavior, which are analyzed by the emotion engine to output the user's emotional state.

[0597] Step 2:

[0598] Data transmission:

[0599] The collected 5W1H data and emotion data are encrypted by the device. The device then sends this encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is a set of emotion and 5W1H data, and the output is an encrypted data stream.

[0600] Step 3:

[0601] Data reception and validation:

[0602] The server decrypts and verifies the data it receives, which involves the process of ensuring that the data entered is accurate and complete. The input is the encrypted data set, and the output is the decrypted data.

[0603] Step 4:

[0604] Data Retention:

[0605] The server then stores the validated data in a database. This process preserves the integrity of the data and properly categorizes and stores it for future analysis and learning. The input is the decrypted data, and the output is a stored entry in the database.

[0606] Step 5:

[0607] Data Analysis:

[0608] The server analyzes the data stored in the database. This analysis process involves aggregating, classifying, and recognizing patterns from the 5W1H data, and combining it with emotional data to find statistical relationships. The input is the raw data retrieved from the database, and the output is an analysis report or pattern recognition results.

[0609] Step 6:

[0610] Strengthening model training:

[0611] Based on the analysis results, the server enhances the learning of the generative AI model, especially the process of optimizing the model parameters by reflecting the user's emotional data. The input is the analysis report and the existing model parameters, and the output is the optimized generative AI model.

[0612] Step 7:

[0613] Updated model delivery:

[0614] The server delivers a generative AI model incorporating the latest learning results to the device. The device downloads this model and applies the new model the next time the user uses the generative AI. The input is the optimized generative AI model, and the output is the new AI model downloaded to the device.

[0615] Step 8:

[0616] Emotional Adaptation Display:

[0617] When a user uses smart glasses, for example, the system selects the most appropriate information or advertisement based on emotional data and displays it in real time. In this process, an algorithm operates to select specific advertisements or information based on the user's current emotional state. The input is emotional data obtained in real time, and the output is advertisements or information displayed on the user interface.

[0618] For example, when the user smiles, an advertisement is selected based on the prompt, "Generate the best advertisement to display when the user is smiling," and an "ice cream sale advertisement" is displayed on the smart glasses display.

[0619] 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.

[0620] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0621] 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.

[0622] [Third embodiment]

[0623] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0624] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0625] 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).

[0626] 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.

[0627] 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.

[0628] 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).

[0629] 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.

[0630] 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.

[0631] 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.

[0632] 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.

[0633] 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.

[0634] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0635] This invention relates to a system that provides more accurate generative AI services by collecting data on the use of generative AI and enhancing the learning of AI models through analysis. Below, we will explain how this system is implemented by the server, terminal, and user.

[0636] Data collection:

[0637] When a user begins using the generation AI, the device automatically detects the user's actions and begins recording data. The device collects data such as the start and end times of use (when), IP address and GPS information (where), user ID and device ID (who), the type of generation request and specific usage (what), purpose information entered by the user (why), and specific instructions and conditions (how).

[0638] Data transmission:

[0639] The collected data is encrypted by the device and sent using a secure communication protocol to a server, which verifies the received data and stores it in a database.

[0640] Data Analysis:

[0641] The server analyzes the received data using aggregation, classification, and pattern recognition algorithms to identify usage frequency, regional needs, and usage trends by time of day, resulting in a detailed analysis of the use of the Generative AI.

[0642] Strengthening model training:

[0643] The server uses data analysis to enhance the training of the generative AI model. The analysis results are incorporated as feedback into the generative AI model, adjusting the model's parameters. A specialized training dataset is created based on specific usage patterns and requests.

[0644] Updated model delivery:

[0645] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[0646] Offering user incentives:

[0647] The server rewards users who cooperate with the data analysis. The rewards are applied to the user's account, notified via the user's device, and can be exchanged for benefits such as free use of the next service.

[0648] As a concrete example, when a user uses article generation AI to create a "report on environmental issues," the device records detailed usage information and sends it to the server. The server aggregates and analyzes this data, detecting patterns such as "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings." The server then uses this information to strengthen the learning of the generative AI model and prepares a specialized training dataset. The updated model is sent to the device, and when the user requests article generation again, a higher-quality, more specific report is generated more quickly.

[0649] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0650] The processing flow will be explained below.

[0651] Step 1:

[0652] To use the generative AI, the user launches the application on the device. The device detects the user's actions and begins preparing to record data.

[0653] Step 2:

[0654] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically collects the 5W1H data (when, where, who, what, why, and how) along with the request.

[0655] Step 3:

[0656] The 5W1H data collected by the device is encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[0657] Step 4:

[0658] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[0659] Step 5:

[0660] The server aggregates and categorizes the data stored in the database and applies pattern recognition algorithms to analyze the data, identifying patterns such as frequency of use, regional needs, and usage trends over time.

[0661] Step 6:

[0662] The server uses the analysis results to enhance the training of the generative AI model, incorporates the analysis results as feedback to adjust the model parameters, and creates specialized training datasets based on specific usage patterns and requests.

[0663] Step 7:

[0664] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[0665] Step 8:

[0666] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain highly accurate and specific generative results.

[0667] Step 9:

[0668] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[0669] Through this series of steps, we provide a system that can efficiently utilize the collected data and improve the accuracy and sophistication of generative AI.

[0670] Example 1

[0671] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0672] In current generative AI systems, the process for efficiently collecting and analyzing user usage data and enhancing model learning has not been fully established. As a result, there are limitations on the quality of the generative AI services provided by users, making it difficult to provide highly accurate and sophisticated generated content. Furthermore, if appropriate incentives are not provided for users to provide data, data collection itself may be insufficient. Therefore, a system that solves these issues, efficiently collects and analyzes generative AI usage data, and provides high-quality services is needed.

[0673] 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.

[0674] In this invention, the server includes means for collecting data when a user uses the generative AI, means for encrypting the collected data and transmitting it to the server, means for verifying the transmitted data and securely storing it in a database, means for analyzing the stored data and identifying user behavior patterns and needs, means for enhancing the learning of the generative AI model based on the analysis results, means for delivering the updated generative AI model to the terminal, means for providing rewards to the user, and means for collecting 5W1H data (when, where, who, what, why, and how). This makes it possible to efficiently collect and analyze generative AI usage data and provide users with highly accurate and advanced generative AI services.

[0675] "Means of collecting data" refers to the function by which the terminal automatically detects and records information generated when the user uses the artificial intelligence.

[0676] "Means for encrypting data and sending it to a server" refers to a function that uses encryption technology to securely convert data collected by the terminal and send it safely to a server via the Internet.

[0677] "Means for verifying data and safely storing it in a database" is a function in which the server checks the integrity and security of the data received and then stores it in a database for safekeeping.

[0678] "Means for analyzing stored data and identifying user behavior patterns and needs" refers to a function that analyzes collected and stored data using analytical tools and algorithms to identify user behavior and needs.

[0679] "Means to enhance the learning of the generative AI model based on the analysis results" refers to a function that improves the performance of the model by optimizing and re-learning the parameters of the generative AI model using the results of data analysis.

[0680] "Means for delivering updated generative artificial intelligence models to terminals" refers to a function that sends enhanced AI models from a server to terminals, making the new models available on the user's system.

[0681] The "means for providing rewards to users" is a function for providing incentives such as points or benefits to users who cooperate in providing data.

[0682] "Means of collecting 5W1H data (when, where, who, what, why, and how)" is a function that comprehensively collects detailed information about users' use of AI.

[0683] The present invention relates to a data collection and analysis system that utilizes a generative AI model, and is implemented as follows.

[0684] To implement the invention, a user uses a generation AI service. When the user begins using the service, the device automatically detects the user's actions and begins collecting related data. Specifically, the device collects data such as the start and end times of use, IP address, GPS information, user ID, device ID, type of generation request and specific usage content, purpose information entered by the user, and specific instructions and conditions. This process uses common computing devices such as mobile devices and PCs.

[0685] The collected data is encrypted by the device, for example using AES encryption technology, and sent to a server via HTTPS, a secure communication protocol. The server verifies the integrity and security of the received data and stores it in a database, which may use MongoDB, a NoSQL database.

[0686] The server uses the Python programming language and its data analysis libraries (such as Pandas and Scikit-learn) to analyze the stored data. During the analysis process, the data is aggregated and classified to identify user behavior patterns and needs. For example, a usage analysis might reveal that "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings."

[0687] Based on the analysis results, the server enhances the learning of the generative AI model (e.g., GPT-3). It retrains the model using machine learning libraries such as TensorFlow or PyTorch to optimize parameters. A specialized dataset is created and the model is updated based on this.

[0688] The updated generative AI model is sent from the server to the device via a REST API, which the device receives and applies the new model the next time the user uses the generative AI, allowing the user to receive an improved generative AI service.

[0689] Furthermore, the server has the function of rewarding users who cooperate in providing data. Users are given points and other benefits, which are notified via their terminals. Users can exchange these points for benefits such as discounts on their next usage fee.

[0690] Examples of specific prompts include:

[0691] "Write an academic report of no more than 500 words on an environmental issue. Specific instructions: Cite the latest data and propose a solution to the problem."

[0692] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0693] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0694] Step 1:

[0695] The user begins using the AI ​​generation service.

[0696] Specifically, a user launches the app and clicks the "Create a report on environmental issues" button. The input includes the type of request and its purpose. The output is a request sent to the device.

[0697] Step 2:

[0698] The device detects user actions and collects data.

[0699] Specifically, the terminal collects the start time of use, end time, IP address, GPS information, user ID, device ID, type of generated request, specific usage content, purpose information entered by the user, specific instructions and conditions, etc. The input is the user's actions, and the output is the collected data that is stored locally.

[0700] Step 3:

[0701] The data collected by the device is encrypted and sent to the server.

[0702] Specifically, the data is encrypted using AES encryption technology and sent to the server using the HTTPS protocol. The input is the collected data, and the output is the encrypted data sent to the server.

[0703] Step 4:

[0704] The server validates the received data and stores it securely in a database.

[0705] Specifically, it verifies the data integrity and encryption key, and stores the data in a MongoDB database. The input is the encrypted data, and the output is the verified data stored in the database.

[0706] Step 5:

[0707] The server analyzes the stored data.

[0708] Specifically, we use Python scripts and data analysis libraries (Pandas, Scikit-learn) to aggregate and classify data to identify user behavior patterns and needs. The input is data stored in a database, and the output is analyzed data.

[0709] Step 6:

[0710] The server enhances the learning of the generative AI model based on the analysis results.

[0711] Specifically, TensorFlow or PyTorch are used to retrain generative AI models (e.g., GPT-3) and adjust their parameters. The inputs are analysis results and specialized datasets, and the output is an enhanced generative AI model.

[0712] Step 7:

[0713] The updated generative AI model is distributed from the server to the device.

[0714] Specifically, a new generative AI model is sent to the device via a REST API, and the device receives it and stores it in storage. The input is the updated generative AI model, and the output is the new model stored on the device.

[0715] Step 8:

[0716] The server provides a reward to the user who cooperates in providing data.

[0717] Specifically, points are added to the user's account and notified via the terminal. The input is the data provision record, and the output is the points added to the user's account and notification.

[0718] In this way, through step-by-step processing and specific operations, generative AI services can be implemented efficiently and to a high degree.

[0719] (Application example 1)

[0720] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0721] Currently, the accuracy of customer service and product recommendations in brick-and-mortar stores is insufficient, making it difficult to improve user satisfaction. Furthermore, there is no system in place to effectively collect and analyze customer behavior data and use the results to enhance the model learning of generative artificial intelligence (generative AI). This makes it difficult to achieve highly accurate customer support and product recommendations using generative AI.

[0722] 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.

[0723] In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for enhancing model learning of the generative AI based on the analysis results, means for distributing the updated generative AI model to the terminal, means for collecting data based on customer behavior in a physical store, and means for making optimal product recommendations based on customer behavior information, thereby enabling highly accurate customer support and product recommendations in physical stores.

[0724] "Generative AI" refers to AI that has the ability to automatically generate natural language sentences, images, etc. based on data.

[0725] "Data collection means" refers to devices or software used to collect information about user behavior and events that occur.

[0726] "Means for sending collected data to a server" refers to a mechanism for transferring collected data to a server via the Internet or a dedicated communication protocol.

[0727] "Means for analyzing transmitted data" refers to software that processes the data received by the server, such as classifying, extracting, and recognizing patterns.

[0728] "Means for enhancing model learning" refers to a process for adjusting the parameters of an artificial intelligence model based on the analysis results to improve performance and accuracy.

[0729] The "means for delivering an updated generative AI model to a terminal" is a mechanism for downloading an AI model updated on a server to a user's terminal.

[0730] "Means for collecting data based on customer behavior within a physical store" refers to devices and systems for collecting information such as customer movements, behavior, and purchase history within a physical store.

[0731] "Means for making optimal product recommendations based on customer behavioral information" refers to a system that utilizes collected customer behavioral data to suggest products and services that are suitable for customers.

[0732] This invention provides a generative AI system for providing highly accurate customer service and product recommendations in brick-and-mortar stores. Specifically, this system collects and analyzes customer behavior data and uses the results to strengthen generative AI models.

[0733] Data collection

[0734] Devices (e.g., smart glasses, smartphones) monitor customer behavior in physical stores and collect data, including customer behavior logs (e.g., product searches, purchase history), location information, and usage time.

[0735] Data transmission

[0736] The collected data is then sent from the device to the server, where it is encrypted using the Fernet library and transmitted over a secure communication protocol.

[0737] Data analysis

[0738] The server analyzes the received data, which includes data classification, extraction, and pattern recognition. Analysis identifies customer needs and behavioral patterns, providing insights for optimal product recommendations.

[0739] Enhanced model learning

[0740] The server then uses the analysis results to enhance the generative AI model, adjusting the model's parameters to improve accuracy in the next recommendation or customer support, for example, by adjusting parameters based on trends in a particular time period or region.

[0741] Delivery of updated models

[0742] The updated generative AI model is sent from the server to the device, which then downloads the new model and applies it the next time it is used, ensuring highly accurate support based on the latest information.

[0743] Providing user incentives

[0744] The server will provide rewards to users who cooperate in providing data. These rewards will be applied to the next purchase as points or discounts. Users will also be notified of these benefits via their devices.

[0745] Hardware and software used

[0746] Hardware: Smart glasses, smartphones

[0747] Software: Fernet library (encryption), pattern recognition algorithms, database

[0748] Specific examples

[0749] If a customer uses smart glasses to search for sneakers in a physical store, the prompt would look like this:

[0750] Request: A customer uses smart glasses to search for sneakers.

[0751] Based on these prompts, the system collects and analyzes customer behavior data to enhance and update the generative AI model. The data based on the prompts is sent to a server, and the analysis results are used to adjust the model's parameters. The updated model is then sent to the device and used for the next customer support call.

[0752] This will improve the accuracy of product recommendations and customer support in physical stores, and is expected to increase customer satisfaction.

[0753] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0754] Step 1:

[0755] When a user searches for a product in a physical store using smart glasses or a smartphone, data on the search action is collected. Specifically, the device records the user's search target (product name, category, etc.), location information, timestamp, etc.

[0756] Input: User search behavior, location, timestamp

[0757] Output: Search behavior data (product name, category, location, timestamp)

[0758] Step 2:

[0759] The device encrypts the collected search behavior data using the Fernet library, and then sends the encrypted data to the server.

[0760] Input: Search behavior data

[0761] Output: Encrypted search behavior data

[0762] Step 3:

[0763] The server decrypts the received encrypted data and performs pre-processing for analysis, such as classifying, organizing, and removing noise from the data.

[0764] Input: Encrypted search behavior data

[0765] Output: Preprocessed data

[0766] Step 4:

[0767] The server analyzes the pre-processed data and uses pattern recognition algorithms to identify customer behavior patterns and needs. The results of this analysis are used as feedback for the generative AI model.

[0768] Input: Preprocessed data

[0769] Output: Analysis results (customer behavior patterns, needs, etc.)

[0770] Step 5:

[0771] Based on the analysis results, the server adjusts the parameters of the generative AI model and strengthens model learning, thereby improving the accuracy of the next recommendation.

[0772] Input: Analysis results

[0773] Output: An updated generative AI model

[0774] Step 6:

[0775] The server delivers the updated generative AI model to the device, which then downloads the new model and uses it for the next customer support or product recommendation.

[0776] Input: Updated generative AI model

[0777] Output: A new generative AI model installed on the device.

[0778] Step 7:

[0779] The server provides incentives to users who cooperate in providing data, such as points or rewards, to their accounts and notifies them via their devices.

[0780] Input: User cooperation data

[0781] Output: Reward notification to user (points, bonuses, etc.)

[0782] 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.

[0783] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[0784] Data collection:

[0785] When a user begins using generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed through the emotion engine.

[0786] Data transmission:

[0787] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS). The server verifies the received data and stores it in a database.

[0788] Data Analysis:

[0789] The server analyzes the received data. This analysis includes aggregation, classification, and pattern recognition algorithms for 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[0790] Strengthening model training:

[0791] Based on the data analysis, the server strengthens the learning of the generative AI model. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for each specific emotional state.

[0792] Updated model delivery:

[0793] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[0794] Offering user incentives:

[0795] The server calculates the reward for users who cooperate with the data analysis. The reward is applied to the user's account and notified via the terminal. As a reward, the user receives benefits such as free use of the next service.

[0796] As a concrete example, when a user uses an article generation AI to create a "report on environmental issues," the device analyzes the user's facial expressions and tone of voice using an emotion engine and collects emotional data. The server analyzes this data to understand the user's emotions toward a specific topic. The server then uses this information to strengthen the learning of the generative AI model and generate more appropriate content that reflects the specific emotions.

[0797] In this way, the present invention efficiently collects and analyzes usage data and emotion data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0798] The processing flow will be explained below.

[0799] Step 1:

[0800] To use generative AI, a user launches an application on their device. The device detects the user's actions and begins recording data and collecting emotional data.

[0801] Step 2:

[0802] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically begins collecting the 5W1H data (when, where, who, what, why, and how) along with the request.

[0803] Step 3:

[0804] The 5W1H data collected by the device and the user's emotional data (e.g., facial expression analysis results and tone of voice analysis results) are encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[0805] Step 4:

[0806] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[0807] Step 5:

[0808] The server analyzes the data stored in the database. The analysis includes aggregation, classification, and pattern recognition algorithms of 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage behavior and their emotional state.

[0809] Step 6:

[0810] The server uses the analysis results to enhance the learning of the generative AI model, incorporates the analysis results as feedback to adjust the model's parameters, and creates specialized training datasets tailored to specific usage patterns and emotional states.

[0811] Step 7:

[0812] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[0813] Step 8:

[0814] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain more accurate and specific generative results that also reflect the user's emotional state.

[0815] Step 9:

[0816] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[0817] Through this series of steps, we provide a system that can efficiently utilize the collected emotional data and improve the accuracy and sophistication of the generative AI. As a specific example, when a user uses the article generation AI to create a "report on environmental issues," the device collects the user's facial expressions and tone of voice as emotional data along with usage information and sends it to the server. The server analyzes this data and strengthens the learning of the generative AI model, thereby generating a high-quality report that takes the user's emotions into consideration.

[0818] Example 2

[0819] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0820] Conventional generative AI systems only collect data when users use the AI, and do not consider user emotional data, which limits the accuracy of the generated content. In addition, there is a lack of feedback and incentives for users, making it difficult to effectively strengthen model learning.

[0821] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for strengthening the model learning of the generative AI based on the analysis results, means for delivering the updated generative AI model to the terminal, means for analyzing user emotion data, means for encrypting and transmitting data including the emotion data, and means for collecting 5W1H data. This enables more effective strengthening of the model learning of the generative AI by generating content with high accuracy that reflects the user's emotion data and providing user incentives.

[0822] "User" refers to an individual or organization that utilizes the System to use Generative Artificial Intelligence.

[0823] "Generative AI" refers to an AI model that generates content such as text and images based on given data and prompts.

[0824] "Data collection means" refers to the technology and devices used to collect user usage data, 5W1H data, and emotional data.

[0825] "Server" refers to the computer system that receives, stores, and analyzes collected data to enhance the learning of the generative AI model.

[0826] "Encryption means" refers to the technology or device that encrypts collected data for secure transmission.

[0827] "Emotional data" refers to emotional information analyzed from the user's facial expressions and tone of voice collected through the device's built-in camera and microphone, or external devices.

[0828] "Emotion engine" refers to software or hardware that analyzes collected emotion data and recognizes the user's emotional state.

[0829] "5W1H data" refers to data that includes information on when, where, who, what, why, and how.

[0830] "Reward means" refers to the technology or mechanism that provides rewards to users who cooperate with data analysis.

[0831] A "generative AI model" refers to an artificial intelligence model that generates content such as text and images based on user input data and analysis results.

[0832] "Model learning enhancement means" refers to the techniques and processes that improve and update the generative AI model based on received data and analysis results.

[0833] "Terminal" refers to a device or equipment used by a user, and may include sensors such as a camera or microphone.

[0834] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[0835] First, when a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). This emotional data is acquired using the device's built-in camera and microphone, or an external device. The collected emotional data is analyzed by an emotion engine running on the device. This analysis identifies the user's emotional state.

[0836] The collected 5W1H data and emotion data are encrypted by the device and sent to a server using a secure communication protocol (e.g., HTTPS), where the server verifies the received data and stores it in a database.

[0837] The server then analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state. Specifically, detailed analysis results can be obtained, such as frequency of use, needs by region, and usage trends by time of day.

[0838] The server strengthens the learning of the generative AI model based on these analysis results. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's response and generated content. For example, when a user feels "concerned," the generative AI can adjust its content to provide more reassuring content. This creates a specialized learning dataset tailored to a specific emotional state.

[0839] The updated generative AI model is sent from the server to the device, which then downloads the updated AI model and applies it the next time the user uses the generative AI, providing a more accurate service.

[0840] In addition, for users who cooperate with data analysis, the server calculates a reward and applies it to the user's account. This reward is notified to the user via their device, and the user can receive benefits such as free use of the service next time.

[0841] A specific example is when a user uses an article generation AI to create a "report on environmental issues." In this case, when the user begins working, the device detects their usage and collects information such as the start time and location. At the same time, the emotion engine analyzes the user's facial expressions and tone of voice via the camera and microphone, collecting emotional data. The server analyzes this data to understand how the user feels about a particular topic. Based on the results of this analysis, the generative AI model strengthens its learning so that it can generate appropriate content that reflects specific emotions.

[0842] Here is an example prompt:

[0843] "We're using generative AI to create a report on environmental issues. We've analyzed the user sentiments and found that 'interesting' and 'concerned' are the two most prominent. Based on this, we'd like you to generate a bullet point summary that reflects those sentiments."

[0844] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0845] Step 1:

[0846] The user begins using the generation AI. The device detects the user's actions. The input is the user's action (for example, launching the article generation AI), and the output is data such as the time and location of use. Specifically, the device records the user's action log.

[0847] Step 2:

[0848] The device collects 5W1H data and emotional data. The input is the user's usage situation (when, where, who, what, why, how) as well as the user's facial expression and tone of voice, and the output is analyzed emotional data and 5W1H data. Specifically, the device uses the built-in camera and microphone to collect data on the user's facial expression and voice, which is then analyzed by the emotion engine.

[0849] Step 3:

[0850] The device encrypts the data collected and analyzed. The input is 5W1H data and analyzed emotional data, and the output is encrypted data. Specifically, the encryption module in the device encrypts the data and converts it into a secure format.

[0851] Step 4:

[0852] The terminal sends encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is the encrypted data, and the output is the completion of data transmission to the server. Specifically, the terminal's communication module sends the data to the server.

[0853] Step 5:

[0854] The server verifies the data it receives and stores it in a database. The input is the encrypted data sent from the device, and the output is data in an analyzable format stored in the database. The specific operation is that the server checks the integrity of the data and stores it in the database.

[0855] Step 6:

[0856] The server analyzes the stored data. The input is the 5W1H data and emotional data stored in the database, and the output is an analysis result that clarifies the relationship between the user's usage and emotional state. Specifically, the server runs aggregation, classification, and pattern recognition algorithms to analyze the data.

[0857] Step 7:

[0858] The server strengthens the learning of the generative AI model based on the analysis results. The input is the analyzed data on the user's usage and emotional state, and the output is an improved generative AI model. Specifically, the server adjusts the parameters of the generative AI model and updates the model's learning.

[0859] Step 8:

[0860] The server delivers the updated generative AI model to the device. The input is the updated generative AI model, and the output is the completion of model delivery to the device. Specifically, the server sends the new generative AI model to the device, which receives and stores it.

[0861] Step 9:

[0862] The server calculates rewards for users who cooperate with data analysis and notifies the users via their devices. The input is the user's cooperation data, and the output is reward information applied to the user's account. Specifically, the server calculates the reward and notifies the result to the user's device.

[0863] Step 10:

[0864] The device downloads the new generative AI model and applies it the next time the user uses the generative AI. The input is the delivered generative AI model, and the output is content generation using the updated generative AI. Specifically, the device saves the new model and applies it the next time the user uses the generative AI.

[0865] Through the above processing steps, we have created a system that uses user emotion data to improve the accuracy of generative AI models.

[0866] (Application example 2)

[0867] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0868] The challenge is to provide more accurate services by efficiently utilizing user emotional data in model training for generative AI. It is also desirable to improve user satisfaction and increase the utilization rate of services by dynamically providing optimal information according to the user's emotional state.

[0869] The specific processing by the specific processing unit 290 of the data processing device 12 in application example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses a generative AI, means for transmitting the collected data and the user's emotional data to the server, means for analyzing the transmitted data and emotional data, means for strengthening model learning of the generative AI based on the analysis results, means for delivering an updated generative AI model to the terminal, and means for selecting and displaying information based on the user's emotional data. This enables a highly accurate generative AI service that reflects the user's emotions.

[0870] "Means for collecting data when a user uses a generative artificial intelligence" refers to a device or software that has the function of recording the actions and inputs a user makes when using a generative artificial intelligence system.

[0871] The "means for transmitting collected data and user emotional data to a server" refers to a device or software that has the function of sending the acquired user usage data and emotional state data to a server using a secure communication protocol.

[0872] "Means for analyzing transmitted data and emotional data" refers to algorithms and software that classify and aggregate the data received by the server and perform detailed analysis, including of the user's emotional state.

[0873] The "means for strengthening the model learning of generative AI based on the analysis results" is a system that has the function of adjusting the parameters of the generative AI model and increasing its adaptability by using user data and emotional data obtained through the analysis.

[0874] "Means for delivering updated generative AI models to terminals" refers to a system that has the function of sending a generative AI model that reflects the latest learning content to a terminal, allowing the user to use the updated model the next time they use it.

[0875] "Means for selecting and displaying information based on user emotional data" refers to a system that analyzes a user's emotional state data, selects the most appropriate information or advertisements, and displays them on the user's device.

[0876] "5W1H data" is a data format that refers to information on when, where, who, what, why, and how.

[0877] This invention is a system that combines data obtained when a user uses generative artificial intelligence (generative AI) with an emotion engine that recognizes the user's emotions to strengthen the model learning of the generative AI and provide a more accurate service. Below, we will explain how the server, terminal, and user each implement this system.

[0878] Data collection:

[0879] When a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed by the emotion engine. This allows for detailed records of the user's usage and their emotional state at the time.

[0880] Data transmission:

[0881] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS), where it is verified and stored in a database.

[0882] Data Analysis:

[0883] The server analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[0884] Strengthening model training:

[0885] The server strengthens the learning of the generative AI model based on data analysis. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for a specific emotional state.

[0886] Updated model delivery:

[0887] The server delivers the generative AI model that reflects the latest learning content to the device, and the device downloads the new AI model and applies it the next time the user uses the generative AI.

[0888] Emotional Adaptation Display:

[0889] Based on the user's emotional data, the system selects the most appropriate information and advertisements and displays them on devices such as smart glasses, thereby providing the most effective advertisements and information to users in real time.

[0890] For example, if a user is wearing smart glasses and the camera detects the user smiling, the application will interpret this as "happiness." Based on this emotional information, the server will select the most appropriate advertisement, such as an advertisement for "ice cream on sale," and display it on the smart glasses' display. In this way, the optimal advertisement is dynamically provided according to the user's emotional state.

[0891] Example prompt sentence:

[0892] Generate the perfect ad to show when the user is smiling.

[0893] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0894] Step 1:

[0895] Data collection:

[0896] When a user begins using the generative AI, the device automatically starts up and detects the user's actions. Specifically, the device's built-in camera and microphone are used to collect the user's 5W1H data as well as emotional data. The input data is the user's voice, facial expressions, and behavior, which are analyzed by the emotion engine to output the user's emotional state.

[0897] Step 2:

[0898] Data transmission:

[0899] The collected 5W1H data and emotion data are encrypted by the device. The device then sends this encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is a set of emotion and 5W1H data, and the output is an encrypted data stream.

[0900] Step 3:

[0901] Data reception and validation:

[0902] The server decrypts and verifies the data it receives, which involves the process of ensuring that the data entered is accurate and complete. The input is the encrypted data set, and the output is the decrypted data.

[0903] Step 4:

[0904] Data Retention:

[0905] The server then stores the validated data in a database. This process preserves the integrity of the data and properly categorizes and stores it for future analysis and learning. The input is the decrypted data, and the output is a stored entry in the database.

[0906] Step 5:

[0907] Data Analysis:

[0908] The server analyzes the data stored in the database. This analysis process involves aggregating, classifying, and recognizing patterns from the 5W1H data, and combining it with emotional data to find statistical relationships. The input is the raw data retrieved from the database, and the output is an analysis report or pattern recognition results.

[0909] Step 6:

[0910] Strengthening model training:

[0911] Based on the analysis results, the server enhances the learning of the generative AI model, especially the process of optimizing the model parameters by reflecting the user's emotional data. The input is the analysis report and the existing model parameters, and the output is the optimized generative AI model.

[0912] Step 7:

[0913] Updated model delivery:

[0914] The server delivers a generative AI model incorporating the latest learning results to the device. The device downloads this model and applies the new model the next time the user uses the generative AI. The input is the optimized generative AI model, and the output is the new AI model downloaded to the device.

[0915] Step 8:

[0916] Emotional Adaptation Display:

[0917] When a user uses smart glasses, for example, the system selects the most appropriate information or advertisement based on emotional data and displays it in real time. In this process, an algorithm operates to select specific advertisements or information based on the user's current emotional state. The input is emotional data obtained in real time, and the output is advertisements or information displayed on the user interface.

[0918] For example, when the user smiles, an advertisement is selected based on the prompt, "Generate the best advertisement to display when the user is smiling," and an "ice cream sale advertisement" is displayed on the smart glasses display.

[0919] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0920] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0921] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[0922] [Fourth embodiment]

[0923] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0924] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0925] 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).

[0926] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[0927] 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.

[0928] 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).

[0929] 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.

[0930] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[0931] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0932] 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.

[0933] 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.

[0934] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0935] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0936] This invention relates to a system that provides more accurate generative AI services by collecting data on the use of generative AI and enhancing the learning of AI models through analysis. Below, we will explain how this system is implemented by the server, terminal, and user.

[0937] Data collection:

[0938] When a user begins using the generation AI, the device automatically detects the user's actions and begins recording data. The device collects data such as the start and end times of use (when), IP address and GPS information (where), user ID and device ID (who), the type of generation request and specific usage (what), purpose information entered by the user (why), and specific instructions and conditions (how).

[0939] Data transmission:

[0940] The collected data is encrypted by the device and sent using a secure communication protocol to a server, which verifies the received data and stores it in a database.

[0941] Data Analysis:

[0942] The server analyzes the received data using aggregation, classification, and pattern recognition algorithms to identify usage frequency, regional needs, and usage trends by time of day, resulting in a detailed analysis of the use of the Generative AI.

[0943] Strengthening model training:

[0944] The server uses data analysis to enhance the training of the generative AI model. The analysis results are incorporated as feedback into the generative AI model, adjusting the model's parameters. A specialized training dataset is created based on specific usage patterns and requests.

[0945] Updated model delivery:

[0946] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[0947] Offering user incentives:

[0948] The server rewards users who cooperate with the data analysis. The rewards are applied to the user's account, notified via the user's device, and can be exchanged for benefits such as free use of the next service.

[0949] As a concrete example, when a user uses article generation AI to create a "report on environmental issues," the device records detailed usage information and sends it to the server. The server aggregates and analyzes this data, detecting patterns such as "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings." The server then uses this information to strengthen the learning of the generative AI model and prepares a specialized training dataset. The updated model is sent to the device, and when the user requests article generation again, a higher-quality, more specific report is generated more quickly.

[0950] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0951] The processing flow will be explained below.

[0952] Step 1:

[0953] To use the generative AI, the user launches the application on the device. The device detects the user's actions and begins preparing to record data.

[0954] Step 2:

[0955] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically collects the 5W1H data (when, where, who, what, why, and how) along with the request.

[0956] Step 3:

[0957] The 5W1H data collected by the device is encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[0958] Step 4:

[0959] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[0960] Step 5:

[0961] The server aggregates and categorizes the data stored in the database and applies pattern recognition algorithms to analyze the data, identifying patterns such as frequency of use, regional needs, and usage trends over time.

[0962] Step 6:

[0963] The server uses the analysis results to enhance the training of the generative AI model, incorporates the analysis results as feedback to adjust the model parameters, and creates specialized training datasets based on specific usage patterns and requests.

[0964] Step 7:

[0965] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[0966] Step 8:

[0967] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain highly accurate and specific generative results.

[0968] Step 9:

[0969] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[0970] Through this series of steps, we provide a system that can efficiently utilize the collected data and improve the accuracy and sophistication of generative AI.

[0971] Example 1

[0972] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0973] In current generative AI systems, the process for efficiently collecting and analyzing user usage data and enhancing model learning has not been fully established. As a result, there are limitations on the quality of the generative AI services provided by users, making it difficult to provide highly accurate and sophisticated generated content. Furthermore, if appropriate incentives are not provided for users to provide data, data collection itself may be insufficient. Therefore, a system that solves these issues, efficiently collects and analyzes generative AI usage data, and provides high-quality services is needed.

[0974] 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.

[0975] In this invention, the server includes means for collecting data when a user uses the generative AI, means for encrypting the collected data and transmitting it to the server, means for verifying the transmitted data and securely storing it in a database, means for analyzing the stored data and identifying user behavior patterns and needs, means for enhancing the learning of the generative AI model based on the analysis results, means for delivering the updated generative AI model to the terminal, means for providing rewards to the user, and means for collecting 5W1H data (when, where, who, what, why, and how). This makes it possible to efficiently collect and analyze generative AI usage data and provide users with highly accurate and advanced generative AI services.

[0976] "Means of collecting data" refers to the function by which the terminal automatically detects and records information generated when the user uses the artificial intelligence.

[0977] "Means for encrypting data and sending it to a server" refers to a function that uses encryption technology to securely convert data collected by the terminal and send it safely to a server via the Internet.

[0978] "Means for verifying data and safely storing it in a database" is a function in which the server checks the integrity and security of the data received and then stores it in a database for safekeeping.

[0979] "Means for analyzing stored data and identifying user behavior patterns and needs" refers to a function that analyzes collected and stored data using analytical tools and algorithms to identify user behavior and needs.

[0980] "Means to enhance the learning of the generative AI model based on the analysis results" refers to a function that improves the performance of the model by optimizing and re-learning the parameters of the generative AI model using the results of data analysis.

[0981] "Means for delivering updated generative artificial intelligence models to terminals" refers to a function that sends enhanced AI models from a server to terminals, making the new models available on the user's system.

[0982] The "means for providing rewards to users" is a function for providing incentives such as points or benefits to users who cooperate in providing data.

[0983] "Means of collecting 5W1H data (when, where, who, what, why, and how)" is a function that comprehensively collects detailed information about users' use of AI.

[0984] The present invention relates to a data collection and analysis system that utilizes a generative AI model, and is implemented as follows.

[0985] To implement the invention, a user uses a generation AI service. When the user begins using the service, the device automatically detects the user's actions and begins collecting related data. Specifically, the device collects data such as the start and end times of use, IP address, GPS information, user ID, device ID, type of generation request and specific usage content, purpose information entered by the user, and specific instructions and conditions. This process uses common computing devices such as mobile devices and PCs.

[0986] The collected data is encrypted by the device, for example using AES encryption technology, and sent to a server via HTTPS, a secure communication protocol. The server verifies the integrity and security of the received data and stores it in a database, which may use MongoDB, a NoSQL database.

[0987] The server uses the Python programming language and its data analysis libraries (such as Pandas and Scikit-learn) to analyze the stored data. During the analysis process, the data is aggregated and classified to identify user behavior patterns and needs. For example, a usage analysis might reveal that "in Tokyo, there are many requests to generate articles related to environmental issues during weekday mornings."

[0988] Based on the analysis results, the server enhances the learning of the generative AI model (e.g., GPT-3). It retrains the model using machine learning libraries such as TensorFlow or PyTorch to optimize parameters. A specialized dataset is created and the model is updated based on this.

[0989] The updated generative AI model is sent from the server to the device via a REST API, which the device receives and applies the new model the next time the user uses the generative AI, allowing the user to receive an improved generative AI service.

[0990] Furthermore, the server has the function of rewarding users who cooperate in providing data. Users are given points and other benefits, which are notified via their terminals. Users can exchange these points for benefits such as discounts on their next usage fee.

[0991] Examples of specific prompts include:

[0992] "Write an academic report of no more than 500 words on an environmental issue. Specific instructions: Cite the latest data and propose a solution to the problem."

[0993] In this way, the present invention efficiently collects and analyzes usage data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[0994] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0995] Step 1:

[0996] The user begins using the AI ​​generation service.

[0997] Specifically, a user launches the app and clicks the "Create a report on environmental issues" button. The input includes the type of request and its purpose. The output is a request sent to the device.

[0998] Step 2:

[0999] The device detects user actions and collects data.

[1000] Specifically, the terminal collects the start time of use, end time, IP address, GPS information, user ID, device ID, type of generated request, specific usage content, purpose information entered by the user, specific instructions and conditions, etc. The input is the user's actions, and the output is the collected data that is stored locally.

[1001] Step 3:

[1002] The data collected by the device is encrypted and sent to the server.

[1003] Specifically, the data is encrypted using AES encryption technology and sent to the server using the HTTPS protocol. The input is the collected data, and the output is the encrypted data sent to the server.

[1004] Step 4:

[1005] The server validates the received data and stores it securely in a database.

[1006] Specifically, it verifies the data integrity and encryption key, and stores the data in a MongoDB database. The input is the encrypted data, and the output is the verified data stored in the database.

[1007] Step 5:

[1008] The server analyzes the stored data.

[1009] Specifically, we use Python scripts and data analysis libraries (Pandas, Scikit-learn) to aggregate and classify data to identify user behavior patterns and needs. The input is data stored in a database, and the output is analyzed data.

[1010] Step 6:

[1011] The server enhances the learning of the generative AI model based on the analysis results.

[1012] Specifically, TensorFlow or PyTorch are used to retrain generative AI models (e.g., GPT-3) and adjust their parameters. The inputs are analysis results and specialized datasets, and the output is an enhanced generative AI model.

[1013] Step 7:

[1014] The updated generative AI model is distributed from the server to the device.

[1015] Specifically, a new generative AI model is sent to the device via a REST API, and the device receives it and stores it in storage. The input is the updated generative AI model, and the output is the new model stored on the device.

[1016] Step 8:

[1017] The server provides a reward to the user who cooperates in providing data.

[1018] Specifically, points are added to the user's account and notified via the terminal. The input is the data provision record, and the output is the points added to the user's account and notification.

[1019] In this way, through step-by-step processing and specific operations, generative AI services can be implemented efficiently and to a high degree.

[1020] (Application example 1)

[1021] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1022] Currently, the accuracy of customer service and product recommendations in brick-and-mortar stores is insufficient, making it difficult to improve user satisfaction. Furthermore, there is no system in place to effectively collect and analyze customer behavior data and use the results to enhance the model learning of generative artificial intelligence (generative AI). This makes it difficult to achieve highly accurate customer support and product recommendations using generative AI.

[1023] 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.

[1024] In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for enhancing model learning of the generative AI based on the analysis results, means for distributing the updated generative AI model to the terminal, means for collecting data based on customer behavior in a physical store, and means for making optimal product recommendations based on customer behavior information, thereby enabling highly accurate customer support and product recommendations in physical stores.

[1025] "Generative AI" refers to AI that has the ability to automatically generate natural language sentences, images, etc. based on data.

[1026] "Data collection means" refers to devices or software used to collect information about user behavior and events that occur.

[1027] "Means for sending collected data to a server" refers to a mechanism for transferring collected data to a server via the Internet or a dedicated communication protocol.

[1028] "Means for analyzing transmitted data" refers to software that processes the data received by the server, such as classifying, extracting, and recognizing patterns.

[1029] "Means for enhancing model learning" refers to a process for adjusting the parameters of an artificial intelligence model based on the analysis results to improve performance and accuracy.

[1030] The "means for delivering an updated generative AI model to a terminal" is a mechanism for downloading an AI model updated on a server to a user's terminal.

[1031] "Means for collecting data based on customer behavior within a physical store" refers to devices and systems for collecting information such as customer movements, behavior, and purchase history within a physical store.

[1032] "Means for making optimal product recommendations based on customer behavioral information" refers to a system that utilizes collected customer behavioral data to suggest products and services that are suitable for customers.

[1033] This invention provides a generative AI system for providing highly accurate customer service and product recommendations in brick-and-mortar stores. Specifically, this system collects and analyzes customer behavior data and uses the results to strengthen generative AI models.

[1034] Data collection

[1035] Devices (e.g., smart glasses, smartphones) monitor customer behavior in physical stores and collect data, including customer behavior logs (e.g., product searches, purchase history), location information, and usage time.

[1036] Data transmission

[1037] The collected data is then sent from the device to the server, where it is encrypted using the Fernet library and transmitted over a secure communication protocol.

[1038] Data analysis

[1039] The server analyzes the received data, which includes data classification, extraction, and pattern recognition. Analysis identifies customer needs and behavioral patterns, providing insights for optimal product recommendations.

[1040] Enhanced model learning

[1041] The server then uses the analysis results to enhance the generative AI model, adjusting the model's parameters to improve accuracy in the next recommendation or customer support, for example, by adjusting parameters based on trends in a particular time period or region.

[1042] Delivery of updated models

[1043] The updated generative AI model is sent from the server to the device, which then downloads the new model and applies it the next time it is used, ensuring highly accurate support based on the latest information.

[1044] Providing user incentives

[1045] The server will provide rewards to users who cooperate in providing data. These rewards will be applied to the next purchase as points or discounts. Users will also be notified of these benefits via their devices.

[1046] Hardware and software used

[1047] Hardware: Smart glasses, smartphones

[1048] Software: Fernet library (encryption), pattern recognition algorithms, database

[1049] Specific examples

[1050] If a customer uses smart glasses to search for sneakers in a physical store, the prompt would look like this:

[1051] Request: A customer uses smart glasses to search for sneakers.

[1052] Based on these prompts, the system collects and analyzes customer behavior data to enhance and update the generative AI model. The data based on the prompts is sent to a server, and the analysis results are used to adjust the model's parameters. The updated model is then sent to the device and used for the next customer support call.

[1053] This will improve the accuracy of product recommendations and customer support in physical stores, and is expected to increase customer satisfaction.

[1054] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1055] Step 1:

[1056] When a user searches for a product in a physical store using smart glasses or a smartphone, data on the search action is collected. Specifically, the device records the user's search target (product name, category, etc.), location information, timestamp, etc.

[1057] Input: User search behavior, location, timestamp

[1058] Output: Search behavior data (product name, category, location, timestamp)

[1059] Step 2:

[1060] The device encrypts the collected search behavior data using the Fernet library, and then sends the encrypted data to the server.

[1061] Input: Search behavior data

[1062] Output: Encrypted search behavior data

[1063] Step 3:

[1064] The server decrypts the received encrypted data and performs pre-processing for analysis, such as classifying, organizing, and removing noise from the data.

[1065] Input: Encrypted search behavior data

[1066] Output: Preprocessed data

[1067] Step 4:

[1068] The server analyzes the pre-processed data and uses pattern recognition algorithms to identify customer behavior patterns and needs. The results of this analysis are used as feedback for the generative AI model.

[1069] Input: Preprocessed data

[1070] Output: Analysis results (customer behavior patterns, needs, etc.)

[1071] Step 5:

[1072] Based on the analysis results, the server adjusts the parameters of the generative AI model and strengthens model learning, thereby improving the accuracy of the next recommendation.

[1073] Input: Analysis results

[1074] Output: An updated generative AI model

[1075] Step 6:

[1076] The server delivers the updated generative AI model to the device, which then downloads the new model and uses it for the next customer support or product recommendation.

[1077] Input: Updated generative AI model

[1078] Output: A new generative AI model installed on the device.

[1079] Step 7:

[1080] The server provides incentives to users who cooperate in providing data, such as points or rewards, to their accounts and notifies them via their devices.

[1081] Input: User cooperation data

[1082] Output: Reward notification to user (points, bonuses, etc.)

[1083] 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.

[1084] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[1085] Data collection:

[1086] When a user begins using generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed through the emotion engine.

[1087] Data transmission:

[1088] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS). The server verifies the received data and stores it in a database.

[1089] Data Analysis:

[1090] The server analyzes the received data. This analysis includes aggregation, classification, and pattern recognition algorithms for 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[1091] Strengthening model training:

[1092] Based on the data analysis, the server strengthens the learning of the generative AI model. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for each specific emotional state.

[1093] Updated model delivery:

[1094] The updated generative AI model is sent from the server to the device, which then downloads the new AI model and applies it the next time the user uses the generative AI.

[1095] Offering user incentives:

[1096] The server calculates the reward for users who cooperate with the data analysis. The reward is applied to the user's account and notified via the terminal. As a reward, the user receives benefits such as free use of the next service.

[1097] As a concrete example, when a user uses an article generation AI to create a "report on environmental issues," the device analyzes the user's facial expressions and tone of voice using an emotion engine and collects emotional data. The server analyzes this data to understand the user's emotions toward a specific topic. The server then uses this information to strengthen the learning of the generative AI model and generate more appropriate content that reflects the specific emotions.

[1098] In this way, the present invention efficiently collects and analyzes usage data and emotion data of generative AI and strengthens its learning, thereby providing users with highly accurate and advanced generative AI services.

[1099] The processing flow will be explained below.

[1100] Step 1:

[1101] To use generative AI, a user launches an application on their device. The device detects the user's actions and begins recording data and collecting emotional data.

[1102] Step 2:

[1103] The user inputs a specific request (e.g., generate an article) to the AI, and the device automatically begins collecting the 5W1H data (when, where, who, what, why, and how) along with the request.

[1104] Step 3:

[1105] The 5W1H data collected by the device and the user's emotional data (e.g., facial expression analysis results and tone of voice analysis results) are encrypted and sent to the server using a secure communication protocol (e.g., HTTPS).

[1106] Step 4:

[1107] The server validates the data it receives to check for invalid or missing data, and stores the validated data in the database.

[1108] Step 5:

[1109] The server analyzes the data stored in the database. The analysis includes aggregation, classification, and pattern recognition algorithms of 5W1H data, as well as analysis of emotional data. This clarifies the relationship between the user's usage behavior and their emotional state.

[1110] Step 6:

[1111] The server uses the analysis results to enhance the learning of the generative AI model, incorporates the analysis results as feedback to adjust the model's parameters, and creates specialized training datasets tailored to specific usage patterns and emotional states.

[1112] Step 7:

[1113] The server sends a notification to distribute the updated generative AI model to the device. The device receives the notification and downloads the latest generative AI model from the server.

[1114] Step 8:

[1115] When the user uses the generative AI again, the device will apply the updated generative AI model, allowing the user to obtain more accurate and specific generative results that also reflect the user's emotional state.

[1116] Step 9:

[1117] The server calculates rewards for users who cooperate with data analysis. The rewards are applied to the user's account and notified via the user's device. As a reward, the user receives benefits such as free use of the next service.

[1118] Through this series of steps, we provide a system that can efficiently utilize the collected emotional data and improve the accuracy and sophistication of the generative AI. As a specific example, when a user uses the article generation AI to create a "report on environmental issues," the device collects the user's facial expressions and tone of voice as emotional data along with usage information and sends it to the server. The server analyzes this data and strengthens the learning of the generative AI model, thereby generating a high-quality report that takes the user's emotions into consideration.

[1119] Example 2

[1120] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1121] Conventional generative AI systems only collect data when users use the AI, and do not consider user emotional data, which limits the accuracy of the generated content. In addition, there is a lack of feedback and incentives for users, making it difficult to effectively strengthen model learning.

[1122] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses the generative AI, means for transmitting the collected data to the server, means for analyzing the transmitted data, means for strengthening the model learning of the generative AI based on the analysis results, means for delivering the updated generative AI model to the terminal, means for analyzing user emotion data, means for encrypting and transmitting data including the emotion data, and means for collecting 5W1H data. This enables more effective strengthening of the model learning of the generative AI by generating content with high accuracy that reflects the user's emotion data and providing user incentives.

[1123] "User" refers to an individual or organization that utilizes the System to use Generative Artificial Intelligence.

[1124] "Generative AI" refers to an AI model that generates content such as text and images based on given data and prompts.

[1125] "Data collection means" refers to the technology and devices used to collect user usage data, 5W1H data, and emotional data.

[1126] "Server" refers to the computer system that receives, stores, and analyzes collected data to enhance the learning of the generative AI model.

[1127] "Encryption means" refers to the technology or device that encrypts collected data for secure transmission.

[1128] "Emotional data" refers to emotional information analyzed from the user's facial expressions and tone of voice collected through the device's built-in camera and microphone, or external devices.

[1129] "Emotion engine" refers to software or hardware that analyzes collected emotion data and recognizes the user's emotional state.

[1130] "5W1H data" refers to data that includes information on when, where, who, what, why, and how.

[1131] "Reward means" refers to the technology or mechanism that provides rewards to users who cooperate with data analysis.

[1132] A "generative AI model" refers to an artificial intelligence model that generates content such as text and images based on user input data and analysis results.

[1133] "Model learning enhancement means" refers to the techniques and processes that improve and update the generative AI model based on received data and analysis results.

[1134] "Terminal" refers to a device or equipment used by a user, and may include sensors such as a camera or microphone.

[1135] This invention is a system that combines data obtained when a user uses generative AI with an emotion engine that recognizes the user's emotions to strengthen the model learning of generative AI and provide a more accurate service. Below, we will explain how this system is implemented by the server, terminal, and user.

[1136] First, when a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). This emotional data is acquired using the device's built-in camera and microphone, or an external device. The collected emotional data is analyzed by an emotion engine running on the device. This analysis identifies the user's emotional state.

[1137] The collected 5W1H data and emotion data are encrypted by the device and sent to a server using a secure communication protocol (e.g., HTTPS), where the server verifies the received data and stores it in a database.

[1138] The server then analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state. Specifically, detailed analysis results can be obtained, such as frequency of use, needs by region, and usage trends by time of day.

[1139] The server strengthens the learning of the generative AI model based on these analysis results. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's response and generated content. For example, when a user feels "concerned," the generative AI can adjust its content to provide more reassuring content. This creates a specialized learning dataset tailored to a specific emotional state.

[1140] The updated generative AI model is sent from the server to the device, which then downloads the updated AI model and applies it the next time the user uses the generative AI, providing a more accurate service.

[1141] In addition, for users who cooperate with data analysis, the server calculates a reward and applies it to the user's account. This reward is notified to the user via their device, and the user can receive benefits such as free use of the service next time.

[1142] A specific example is when a user uses an article generation AI to create a "report on environmental issues." In this case, when the user begins working, the device detects their usage and collects information such as the start time and location. At the same time, the emotion engine analyzes the user's facial expressions and tone of voice via the camera and microphone, collecting emotional data. The server analyzes this data to understand how the user feels about a particular topic. Based on the results of this analysis, the generative AI model strengthens its learning so that it can generate appropriate content that reflects specific emotions.

[1143] Here is an example prompt:

[1144] "We're using generative AI to create a report on environmental issues. We've analyzed the user sentiments and found that 'interesting' and 'concerned' are the two most prominent. Based on this, we'd like you to generate a bullet point summary that reflects those sentiments."

[1145] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1146] Step 1:

[1147] The user begins using the generation AI. The device detects the user's actions. The input is the user's action (for example, launching the article generation AI), and the output is data such as the time and location of use. Specifically, the device records the user's action log.

[1148] Step 2:

[1149] The device collects 5W1H data and emotional data. The input is the user's usage situation (when, where, who, what, why, how) as well as the user's facial expression and tone of voice, and the output is analyzed emotional data and 5W1H data. Specifically, the device uses the built-in camera and microphone to collect data on the user's facial expression and voice, which is then analyzed by the emotion engine.

[1150] Step 3:

[1151] The device encrypts the data collected and analyzed. The input is 5W1H data and analyzed emotional data, and the output is encrypted data. Specifically, the encryption module in the device encrypts the data and converts it into a secure format.

[1152] Step 4:

[1153] The terminal sends encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is the encrypted data, and the output is the completion of data transmission to the server. Specifically, the terminal's communication module sends the data to the server.

[1154] Step 5:

[1155] The server verifies the data it receives and stores it in a database. The input is the encrypted data sent from the device, and the output is data in an analyzable format stored in the database. The specific operation is that the server checks the integrity of the data and stores it in the database.

[1156] Step 6:

[1157] The server analyzes the stored data. The input is the 5W1H data and emotional data stored in the database, and the output is an analysis result that clarifies the relationship between the user's usage and emotional state. Specifically, the server runs aggregation, classification, and pattern recognition algorithms to analyze the data.

[1158] Step 7:

[1159] The server strengthens the learning of the generative AI model based on the analysis results. The input is the analyzed data on the user's usage and emotional state, and the output is an improved generative AI model. Specifically, the server adjusts the parameters of the generative AI model and updates the model's learning.

[1160] Step 8:

[1161] The server delivers the updated generative AI model to the device. The input is the updated generative AI model, and the output is the completion of model delivery to the device. Specifically, the server sends the new generative AI model to the device, which receives and stores it.

[1162] Step 9:

[1163] The server calculates rewards for users who cooperate with data analysis and notifies the users via their devices. The input is the user's cooperation data, and the output is reward information applied to the user's account. Specifically, the server calculates the reward and notifies the result to the user's device.

[1164] Step 10:

[1165] The device downloads the new generative AI model and applies it the next time the user uses the generative AI. The input is the delivered generative AI model, and the output is content generation using the updated generative AI. Specifically, the device saves the new model and applies it the next time the user uses the generative AI.

[1166] Through the above processing steps, we have created a system that uses user emotion data to improve the accuracy of generative AI models.

[1167] (Application example 2)

[1168] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1169] The challenge is to provide more accurate services by efficiently utilizing user emotional data in model training for generative AI. It is also desirable to improve user satisfaction and increase the utilization rate of services by dynamically providing optimal information according to the user's emotional state.

[1170] The specific processing by the specific processing unit 290 of the data processing device 12 in application example 2 is realized by the following means. In this invention, the server includes means for collecting data when a user uses a generative AI, means for transmitting the collected data and the user's emotional data to the server, means for analyzing the transmitted data and emotional data, means for strengthening model learning of the generative AI based on the analysis results, means for delivering an updated generative AI model to the terminal, and means for selecting and displaying information based on the user's emotional data. This enables a highly accurate generative AI service that reflects the user's emotions.

[1171] "Means for collecting data when a user uses a generative artificial intelligence" refers to a device or software that has the function of recording the actions and inputs a user makes when using a generative artificial intelligence system.

[1172] The "means for transmitting collected data and user emotional data to a server" refers to a device or software that has the function of sending the acquired user usage data and emotional state data to a server using a secure communication protocol.

[1173] "Means for analyzing transmitted data and emotional data" refers to algorithms and software that classify and aggregate the data received by the server and perform detailed analysis, including of the user's emotional state.

[1174] The "means for strengthening the model learning of generative AI based on the analysis results" is a system that has the function of adjusting the parameters of the generative AI model and increasing its adaptability by using user data and emotional data obtained through the analysis.

[1175] "Means for delivering updated generative AI models to terminals" refers to a system that has the function of sending a generative AI model that reflects the latest learning content to a terminal, allowing the user to use the updated model the next time they use it.

[1176] "Means for selecting and displaying information based on user emotional data" refers to a system that analyzes a user's emotional state data, selects the most appropriate information or advertisements, and displays them on the user's device.

[1177] "5W1H data" is a data format that refers to information on when, where, who, what, why, and how.

[1178] This invention is a system that combines data obtained when a user uses generative artificial intelligence (generative AI) with an emotion engine that recognizes the user's emotions to strengthen the model learning of the generative AI and provide a more accurate service. Below, we will explain how the server, terminal, and user each implement this system.

[1179] Data collection:

[1180] When a user begins using the generative AI, the device automatically detects the user's actions and collects the user's emotional data along with 5W1H data (when, where, who, what, why, and how). The emotional data is acquired from the device's built-in camera and microphone or an external device and analyzed by the emotion engine. This allows for detailed records of the user's usage and their emotional state at the time.

[1181] Data transmission:

[1182] The collected data (5W1H data and emotion data) is encrypted by the device and sent to the server using a secure communication protocol (e.g., HTTPS), where it is verified and stored in a database.

[1183] Data Analysis:

[1184] The server analyzes the received data. This analysis involves the aggregation, classification, and pattern recognition algorithms of the 5W1H data, as well as the analysis of emotional data. This clarifies the relationship between the user's usage and emotional state, and provides detailed analysis results such as usage frequency, regional needs, and usage trends by time of day.

[1185] Strengthening model training:

[1186] The server strengthens the learning of the generative AI model based on data analysis. In particular, if the data includes user emotional data, the server uses that information to optimize the generative AI's reactions and generated content. The analysis results are incorporated as feedback into the generative AI model, and the model's parameters are adjusted. A specialized training dataset is created for a specific emotional state.

[1187] Updated model delivery:

[1188] The server delivers the generative AI model that reflects the latest learning content to the device, and the device downloads the new AI model and applies it the next time the user uses the generative AI.

[1189] Emotional Adaptation Display:

[1190] Based on the user's emotional data, the system selects the most appropriate information and advertisements and displays them on devices such as smart glasses, thereby providing the most effective advertisements and information to users in real time.

[1191] For example, if a user is wearing smart glasses and the camera detects the user smiling, the application will interpret this as "happiness." Based on this emotional information, the server will select the most appropriate advertisement, such as an advertisement for "ice cream on sale," and display it on the smart glasses' display. In this way, the optimal advertisement is dynamically provided according to the user's emotional state.

[1192] Example prompt sentence:

[1193] Generate the perfect ad to show when the user is smiling.

[1194] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1195] Step 1:

[1196] Data collection:

[1197] When a user begins using the generative AI, the device automatically starts up and detects the user's actions. Specifically, the device's built-in camera and microphone are used to collect the user's 5W1H data as well as emotional data. The input data is the user's voice, facial expressions, and behavior, which are analyzed by the emotion engine to output the user's emotional state.

[1198] Step 2:

[1199] Data transmission:

[1200] The collected 5W1H data and emotion data are encrypted by the device. The device then sends this encrypted data to the server using a secure communication protocol (e.g., HTTPS). The input is a set of emotion and 5W1H data, and the output is an encrypted data stream.

[1201] Step 3:

[1202] Data reception and validation:

[1203] The server decrypts and verifies the data it receives, which involves the process of ensuring that the data entered is accurate and complete. The input is the encrypted data set, and the output is the decrypted data.

[1204] Step 4:

[1205] Data Retention:

[1206] The server then stores the validated data in a database. This process preserves the integrity of the data and properly categorizes and stores it for future analysis and learning. The input is the decrypted data, and the output is a stored entry in the database.

[1207] Step 5:

[1208] Data Analysis:

[1209] The server analyzes the data stored in the database. This analysis process involves aggregating, classifying, and recognizing patterns from the 5W1H data, and combining it with emotional data to find statistical relationships. The input is the raw data retrieved from the database, and the output is an analysis report or pattern recognition results.

[1210] Step 6:

[1211] Strengthening model training:

[1212] Based on the analysis results, the server enhances the learning of the generative AI model, especially the process of optimizing the model parameters by reflecting the user's emotional data. The input is the analysis report and the existing model parameters, and the output is the optimized generative AI model.

[1213] Step 7:

[1214] Updated model delivery:

[1215] The server delivers a generative AI model incorporating the latest learning results to the device. The device downloads this model and applies the new model the next time the user uses the generative AI. The input is the optimized generative AI model, and the output is the new AI model downloaded to the device.

[1216] Step 8:

[1217] Emotional Adaptation Display:

[1218] When a user uses smart glasses, for example, the system selects the most appropriate information or advertisement based on emotional data and displays it in real time. In this process, an algorithm operates to select specific advertisements or information based on the user's current emotional state. The input is emotional data obtained in real time, and the output is advertisements or information displayed on the user interface.

[1219] For example, when the user smiles, an advertisement is selected based on the prompt, "Generate the best advertisement to display when the user is smiling," and an "ice cream sale advertisement" is displayed on the smart glasses display.

[1220] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1221] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1222] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1223] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1224] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1225] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1226] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1227] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1228] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1229] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1230] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1231] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1232] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1233] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1234] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1235] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1236] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1237] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1238] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1239] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1240] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1241] The following is further disclosed regarding the above embodiment.

[1242] (Claim 1)

[1243] A means for collecting data when a user uses the artificial intelligence;

[1244] means for transmitting the collected data to a server;

[1245] means for analyzing the transmitted data;

[1246] A means for enhancing model learning of the generative artificial intelligence based on the analysis results;

[1247] means for delivering the updated generative artificial intelligence model to the terminal;

[1248] A system including:

[1249] (Claim 2)

[1250] 10. The system of claim 1,

[1251] A system including a means for providing a reward to a user.

[1252] (Claim 3)

[1253] 10. The system of claim 1,

[1254] A system that includes a means of collecting 5W1H (when, where, who, what, why, and how) data.

[1255] "Example 1"

[1256] (Claim 1)

[1257] A means for collecting data when a user uses the artificial intelligence;

[1258] A means for encrypting the collected data and transmitting it to a server;

[1259] A means to validate the submitted data and store it securely in a database;

[1260] means for analyzing the stored data to identify user behavior patterns and needs;

[1261] a means for enhancing the learning of the generative artificial intelligence model based on the analysis results;

[1262] means for delivering the updated generative artificial intelligence model to the terminal;

[1263] A system including:

[1264] (Claim 2)

[1265] 10. The system of claim 1, further comprising means for providing a reward to the user.

[1266] (Claim 3)

[1267] 10. The system of claim 1, further comprising means for collecting 5W1H data (when, where, who, what, why, and how).

[1268] "Application Example 1"

[1269] (Claim 1)

[1270] A means for collecting data when a user uses the artificial intelligence;

[1271] means for transmitting the collected data to a server;

[1272] means for analyzing the transmitted data;

[1273] A means for enhancing model learning of the generative artificial intelligence based on the analysis results;

[1274] means for delivering the updated generative artificial intelligence model to the terminal;

[1275] A means of collecting data based on customer behavior in physical stores;

[1276] A means for making optimal product recommendations based on customer behavior information;

[1277] A system including:

[1278] (Claim 2)

[1279] A system including a means for providing a reward to a user.

[1280] (Claim 3)

[1281] 10. The system of claim 1, further comprising means for collecting 5W1H (when, where, who, what, why, and how) data.

[1282] "Example 2: Combining Emotion Engines"

[1283] (Claim 1)

[1284] A means for collecting data when a user uses the artificial intelligence;

[1285] means for transmitting the collected data to a server;

[1286] means for analyzing the transmitted data;

[1287] A means for enhancing model learning of the generative artificial intelligence based on the analysis results;

[1288] means for delivering the updated generative artificial intelligence model to the terminal;

[1289] means for analyzing user emotion data;

[1290] means for encrypting and transmitting data including emotion data;

[1291] A means of collecting 5W1H data,

[1292] A system including:

[1293] (Claim 2)

[1294] 10. The system of claim 1, wherein the system provides a reward to the user.

[1295] (Claim 3)

[1296] 10. The system of claim 1, wherein the system analyzes user sentiment regarding a particular topic.

[1297] "Application example 2 when combining emotion engines"

[1298] (Claim 1)

[1299] A means for collecting data when a user uses the artificial intelligence;

[1300] means for transmitting the collected data and the user's emotion data to a server;

[1301] means for analyzing the transmitted data and emotion data;

[1302] A means for enhancing model learning of the generative artificial intelligence based on the analysis results;

[1303] means for delivering the updated generative artificial intelligence model to the terminal;

[1304] means for selecting and displaying information based on the user's emotion data;

[1305] A system including:

[1306] (Claim 2)

[1307] 10. The system of claim 1, further comprising means for providing a reward to the user.

[1308] (Claim 3)

[1309] 10. The system of claim 1, further comprising means for collecting 5W1H (when, where, who, what, why, and how) data. [Explanation of symbols]

[1310] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for collecting data when a user uses the artificial intelligence; means for transmitting the collected data to a server; means for analyzing the transmitted data; A means for enhancing model learning of the generative artificial intelligence based on the analysis results; means for delivering the updated generative artificial intelligence model to the terminal; A system including:

2. 10. The system of claim 1, A system including a means for providing a reward to a user.

3. 10. The system of claim 1, A system that includes a means of collecting 5W1H (when, where, who, what, why, and how) data.

Images