system

Abstract

Provide a system. 【Solution means】 Means for collecting user operation data, Means for sending the collected operation data to the server, Means for saving and initially processing the sent operation data, Means for training an artificial intelligence model that performs learning using the saved and initially processed operation data, Means for generating a product proposal based on the trained artificial intelligence model, Means for displaying the generated product proposal to the user, Means for collecting response data from the user and sending it to the server, A system including means for evaluating the collected response data and updating the artificial intelligence model.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, 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

Summary of the Invention

Problems to be Solved by the Invention

[0004] In modern e-commerce, in order to enhance the purchasing desire of users, personalized product recommendations based on individual user preferences and behavior histories are important. However, in conventional systems, it is difficult to effectively utilize diverse user behavior data to make optimal product recommendations, and there is a problem of low accuracy of the recommendations. Furthermore, since there is a lack of a mechanism to quickly improve the content of the recommendations by reflecting feedback from users, there is a problem that the improvement of the user experience is not sufficiently achieved.

Means for Solving the Problems

[0005] The present invention provides means for collecting user behavior data, means for transmitting the collected behavior data to a server, and means for storing and initial processing the transmitted behavior data. Furthermore, it provides means for training an artificial intelligence model that learns using the stored and initial processed behavior data. The invention also includes means for generating product suggestions based on the trained artificial intelligence model and means for displaying the generated product suggestions to the user. In addition, a system including means for collecting response data from the user and transmitting it to a server, and means for evaluating the collected response data and updating the artificial intelligence model, can realize highly accurate product suggestions based on individual user preferences and improve the user experience.

[0006] "User activity data" refers to records of a user's online behavior, including purchase history, browsing history, and feedback information.

[0007] A "server" is a computer system that receives, stores, and processes data sent from a user terminal.

[0008] "Means of collection" refers to a device or software that has the function of acquiring user behavior data, converting it into a data format, and transmitting it to a server.

[0009] "Means for saving and initial processing" refers to the process of storing received data in a database on the server and correcting inconsistencies and omissions.

[0010] "Means for training an artificial intelligence model that performs learning" refers to the technologies and algorithms used to learn user preferences and behavioral patterns based on collected behavioral data.

[0011] "Means for generating product suggestions" refers to the process of creating appropriate product lists and recommended products for users based on the results of a trained artificial intelligence model.

[0012] "Means of display" refers to interfaces or devices that notify the user's terminal of the generated product suggestions, allowing the user to view them.

[0013] "Response data" refers to records of actions taken by users and feedback information regarding suggested products.

[0014] The "means of evaluation" refer to the process of analyzing the collected response data and providing the results as feedback to the artificial intelligence model.

[0015] "Methods for updating artificial intelligence models" refer to the process of improving accuracy by retraining existing artificial intelligence models based on evaluated response data. [Brief explanation of the drawing]

[0016] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when the emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when the emotion engine is combined.

Mode for Carrying Out the Invention

[0017] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described according to the accompanying drawings.

[0018] First, the terms used in the following description will be described.

[0019] In the following embodiments, a processor with a reference numeral (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0020] In the following embodiments, a RAM (Random Access Memory) with a reference numeral is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0021] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0022] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0024] [First Embodiment]

[0025] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0026] As shown in Figure 1, the 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.

[0027] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0029] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and 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.

[0030] 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 perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0032] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] As shown in Figure 2, in the data processing device 12, a specific processing 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" related to the technology of this 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 according to the specific processing program 56 executed on the RAM 30.

[0034] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0036] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0037] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components.

[0038] 1. Collecting user behavior data

[0039] User terminal operation

[0040] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[0041] 2. Data transmission and storage

[0042] User terminal operation

[0043] The collected behavioral data is converted into a format and sent to the server. For example, it may include information such as the frequency of access to a specific product category and the frequency of purchases.

[0044] Server operation

[0045] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[0046] 3. Training of artificial intelligence models

[0047] Server operation

[0048] Using cleansed data, an artificial intelligence model is trained to learn user preferences and behavioral patterns. For example, past purchase and browsing history is used to learn features that indicate a user's interest in outdoor activities.

[0049] 4. Generating Product Proposals

[0050] Server operation

[0051] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior and current trends. For example, as summer approaches, recommendations for trail running shoes and camping gear will be made.

[0052] User terminal operation

[0053] The user terminal receives the suggestion data sent from the server and displays it through the user interface.

[0054] 5. Collecting user responses and feedback

[0055] User actions

[0056] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[0057] User terminal operation

[0058] This response data is collected and sent to the server.

[0059] 6. Evaluating responses and updating artificial intelligence models

[0060] Server operation

[0061] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, the characteristics of products that users have left particularly positive reviews on are taken into consideration and reflected in subsequent recommendations.

[0062] Specific example

[0063] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed to them. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations. In this way, the entire system is designed to continuously improve itself to enhance the user experience.

[0064] The following describes the processing flow.

[0065] Step 1:

[0066] User actions

[0067] When a user accesses an e-commerce site and browses products, information about the viewed products is recorded in real time. This includes actions such as the user navigating to the product details page, checking reviews, and adding the product to their cart. Furthermore, if the user completes a purchase, that purchase information is also recorded.

[0068] Step 2:

[0069] User terminal operation

[0070] The user's terminal converts collected browsing history, purchase history, and review information into a specified data format. This data includes details such as product ID, viewing time, number of purchases, and review content. The converted data is then sent to the server.

[0071] Step 3:

[0072] Server operation

[0073] The server receives data sent from the user's terminal. The received data is saved to the database in real time. During the saving process, data cleansing is performed. This cleansing process includes supplementing missing data and correcting inconsistent data.

[0074] Step 4:

[0075] Server operation

[0076] The AI ​​model is trained using the cleansed data. This AI model extracts features from user behavior data and learns user preferences and patterns. For example, it learns that a user likes "outdoor activities" based on their past purchase and browsing history.

[0077] Step 5:

[0078] Server operation

[0079] Using a trained AI model, product recommendations are generated for the user. These recommendations combine the user's past behavior with current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[0080] Step 6:

[0081] User terminal operation

[0082] The system receives product suggestion data sent from the server and converts it into a display format. The converted product suggestion list is then displayed to the user through an easy-to-view interface. The user can then browse the suggested products and decide whether or not to purchase them.

[0083] Step 7:

[0084] User actions

[0085] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. Further data is generated when users provide reviews and feedback after purchase.

[0086] Step 8:

[0087] User terminal operation

[0088] The system collects data on user purchases, reviews, and feedback. This data is then converted back into a specified format and sent to the server. This data includes product IDs, rating scores, and review comments.

[0089] Step 9:

[0090] Server operation

[0091] The server analyzes the response data received from the user's terminal. This analysis helps to understand the user's latest preferences and behavioral patterns. For example, if a user gives a high rating to a particular brand of camping equipment, that information will be used as a basis for the analysis.

[0092] Step 10:

[0093] Server operation

[0094] The AI ​​model is retrained based on the new response data. This updates the AI ​​model to reflect the latest data, enabling more accurate product recommendations in the future. New features and patterns are learned, continuously improving the model's performance.

[0095] In this way, the program aims to improve the user experience by collecting user behavior data in real time and using AI to analyze and make suggestions.

[0096] (Example 1)

[0097] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0098] Traditional e-commerce systems have struggled to provide optimal product recommendations based on individual user preferences and behavior. In particular, the lack of mechanisms to collect and rapidly analyze user behavior data in real time for product recommendations led to a decline in the quality of the user experience. Furthermore, insufficient data cleansing processes to ensure the accuracy of collected data meant that improving the accuracy of artificial intelligence models remained a challenge.

[0099] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0100] In this invention, the server includes means for storing and cleaning collected behavioral data, means for training an artificial intelligence model that learns using the stored and cleansed behavioral data, and means for generating product suggestions based on the trained artificial intelligence model. This makes it possible to provide more accurate product suggestions in real time based on the behavioral data of individual users.

[0101] "User activity data" refers to information about all actions and behaviors that a user performs on a website or application. This includes browsing history, purchase history, feedback, and access frequency.

[0102] A "server" refers to a computer system that receives, stores, and processes data sent from user terminals, trains artificial intelligence models, and generates product suggestions.

[0103] "Means of collection" refers to methods and devices for capturing and recording user behavior data. This includes data capture by browser scripts and mobile applications.

[0104] "Means of transmission" refers to the means of transferring collected behavioral data from the user's terminal to the server. This includes HTTP requests and API calls.

[0105] "Means for saving and cleansing data" refers to means of saving the transmitted data to a database and performing data cleansing such as imputing missing values ​​and checking data types.

[0106] An "artificial intelligence model" refers to an algorithm or system that learns user behavior patterns and preferences through machine learning or deep learning based on data.

[0107] "Training methods" refer to methods and devices for training artificial intelligence models using cleansed data. This includes machine learning frameworks (e.g., Tensorflow® and PyTorch).

[0108] "Methods for generating product suggestions" refers to methods that use a trained artificial intelligence model to select the most suitable products for the user and create suggestions.

[0109] "Means of displaying through a user interface" refers to methods and devices for visually presenting generated product suggestions to the user. This includes user interfaces for web pages and mobile applications.

[0110] "Response data" refers to data such as user feedback, purchase history, and ratings regarding suggested products.

[0111] "Means of analysis and updating" refers to methods of analyzing collected response data and retraining an artificial intelligence model based on the results to improve its accuracy.

[0112] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components and processing steps.

[0113] User behavior data collection

[0114] User actions

[0115] User behavioral data is collected when users browse, purchase, and provide feedback on products on e-commerce sites. If a user frequently browses a specific product category (e.g., sporting goods) and purchases related products, that data is collected in real time. Browser scripts and mobile applications are used for this purpose.

[0116] Sending and storing data

[0117] User terminal operation

[0118] The collected behavioral data is converted to JSON format and sent to the server via an HTTP request. For example, it may include logs of when a user frequently viewed a particular product category and information about products they purchased.

[0119] Server operation

[0120] The server stores the received data in a database such as MongoDB or MySQL®. The stored data is then cleansed to maintain data accuracy, including imputation of missing values ​​and data type checks.

[0121] Training of artificial intelligence models

[0122] Server operation

[0123] Based on the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow and PyTorch. For example, features are extracted from a user's past purchase and browsing history to determine if the user is interested in outdoor activities.

[0124] Product Proposal Generation

[0125] Server operation

[0126] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior data and current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[0127] User terminal operation

[0128] The user terminal receives the suggestion data sent from the server and displays it through the user interface. For example, a product list might be displayed in the format, "Here are some camping items we recommend for you."

[0129] Collecting user responses and feedback

[0130] User actions

[0131] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[0132] User terminal operation

[0133] This response data is collected and sent to the server. For example, it may include data such as "I purchased this product" or review information such as "I gave it a 5-star rating."

[0134] Response evaluation and AI model updates

[0135] Server operation

[0136] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, it learns the characteristics of products that users have left particularly positive reviews on and incorporates this into subsequent recommendations.

[0137] Specific example

[0138] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed through the user interface. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations.

[0139] Examples of prompts to input into a generative AI model

[0140] "Generate suggestions recommending new camping equipment, scheduled for release next month, to users who have previously purchased outdoor gear."

[0141] "Create a list of optimal product suggestions based on the sports equipment that users frequently view."

[0142] Thus, the present invention is a system that uses user behavior data to make optimal product recommendations and improve the user experience.

[0143] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0144] Step 1:

[0145] User behavior data collection

[0146] User actions

[0147] Users browse e-commerce sites, purchase products, and provide feedback. This behavioral data is captured by each user's device (browser or mobile application). For example, if a user browses products in the "camping equipment" category for more than 10 minutes and adds a specific product to their cart, that information will be collected.

[0148] input

[0149] User browsing history, purchase history, and feedback information

[0150] output

[0151] Local cache of collected behavioral data

[0152] Step 2:

[0153] Sending data

[0154] User terminal operation

[0155] Behavioral data stored in the local cache is converted to JSON format and sent to the server using an HTTP request. For example, JSON data containing access frequency and purchase information for a specific category is generated and sent to the server.

[0156] input

[0157] Local cache behavioral data

[0158] output

[0159] Action data in JSON format sent to the server

[0160] Step 3:

[0161] Data storage and cleansing process

[0162] Server operation

[0163] The server saves the received JSON data to a database (e.g., MongoDB or MySQL). Afterward, a cleansing process is performed, including data imputation and data type checks. For example, if there are missing values ​​in the purchase history, they will be appropriately imputed.

[0164] input

[0165] Behavioral data in JSON format that was sent.

[0166] output

[0167] Cleansed database behavioral data

[0168] Step 4:

[0169] Training of artificial intelligence models

[0170] Server operation

[0171] Using the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow or PyTorch. Here, behavioral patterns and preferences are learned based on the user's past purchase and browsing history. For example, the pattern of a user frequently purchasing "outdoor equipment" is learned.

[0172] input

[0173] Cleansed behavioral data

[0174] output

[0175] Trained artificial intelligence model

[0176] Step 5:

[0177] Product Proposal Generation

[0178] Server operation

[0179] Using a trained model, the system generates optimal product recommendations based on current user data (such as recent browsing history and purchase information). For example, it might suggest new camping equipment to a user interested in "outdoor goods."

[0180] input

[0181] Current user data, trained artificial intelligence model

[0182] output

[0183] Generated product suggestion data (e.g., JSON format)

[0184] Step 6:

[0185] Product suggestion display

[0186] User terminal operation

[0187] The system analyzes product suggestion data received from the server and displays it through the user interface. For example, it can display a list of "recommended camping gear" on a webpage or mobile app screen.

[0188] input

[0189] Product proposal data sent from the server

[0190] output

[0191] Product suggestions displayed to the user

[0192] Step 7:

[0193] Collecting user responses

[0194] User actions

[0195] Users purchase suggested products, write reviews, and provide feedback. These actions are captured as response data.

[0196] input

[0197] User actions (purchase, review, feedback)

[0198] output

[0199] Local cache of collected response data

[0200] Step 8:

[0201] Sending and analyzing response data

[0202] User terminal operation

[0203] The collected response data is sent to the server. The data is then converted back to JSON format and sent via an HTTP request.

[0204] Server operation

[0205] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, improving the accuracy of future suggestions. For example, the characteristics of products that received high ratings can be reflected in future suggestions.

[0206] input

[0207] Response data sent from the user terminal

[0208] output

[0209] AI model updated based on feedback

[0210] (Application Example 1)

[0211] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0212] Traditional e-commerce systems collect user behavior data to suggest products, but they have shortcomings in the accuracy and timeliness of these suggestions. Furthermore, it was difficult to evaluate in real time how beneficial the suggested products were to the user and incorporate that feedback into future suggestions. Additionally, the lack of smooth integration between product suggestions and the purchase process within the virtual environment resulted in a limited user experience.

[0213] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0214] In this invention, the server includes means for collecting user behavior data, means for transmitting the collected behavior data to the server, means for storing and initial processing the transmitted behavior data, means for training an artificial intelligence model that learns using the stored and initial processed behavior data, means for generating product suggestions based on the trained artificial intelligence model, means for displaying the generated product suggestions to the user, means for collecting and transmitting response data from the user to the server, means for evaluating the collected response data and updating the artificial intelligence model, and means for displaying product suggestions via a visual device used by the user in a virtual environment, enabling the selection and purchase of suggested products. This enables highly accurate and timely product suggestions based on user behavior data, and provides a seamless product selection and purchase experience in a virtual environment.

[0215] "Action data" refers to data about a user's actions and choices within an online or virtual environment, such as browsing history, purchase history, and feedback information.

[0216] A "server" refers to a computer system that receives, stores, and processes operational data over a network, and is used for training artificial intelligence models and generating product suggestions.

[0217] An "artificial intelligence model" refers to software that includes algorithms that analyze user preferences and behavior based on trained data and provide optimal product recommendations.

[0218] "Product recommendations" refer to information that indicates products or services recommended to the user, based on collected behavioral data and analysis results from artificial intelligence models.

[0219] "Visual devices" refer to devices that users wear to display information in a virtual environment, such as smart glasses and head-mounted displays.

[0220] "Response data" refers to information that shows a user's reaction and feedback to a suggested product, such as whether they purchased the product or wrote an article or review.

[0221] "Training" refers to the process of using collected and initially processed behavioral data to train an artificial intelligence model, thereby improving its prediction and suggestion accuracy.

[0222] A "virtual environment" refers to a virtual commercial space or store that a user experiences through a visual device using virtual reality technology.

[0223] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system improves the user experience through the following steps.

[0224] The system first collects user behavior data. This is done using visual devices such as smart glasses or head-mounted displays. These devices collect the user's gaze and gaze duration in real time, recording what the user is looking at.

[0225] The collected motion data is sent from the visual device to the server. The server stores the received data in a database and performs initial processing. This initial processing includes data cleansing, for example, removing noise from the gaze data and extracting accurate motion patterns.

[0226] Next, the server uses the cleansed data to train an artificial intelligence model. This training utilizes past purchase and browsing history. The trained AI model learns the user's preferences and behavioral patterns and generates optimal product recommendations.

[0227] Once product suggestions are generated, the server sends them to the user's visual device for display. The user can then review, select, and purchase the suggested products through their visual device. This allows the user to have a seamless product selection and purchase experience within the virtual environment.

[0228] Additionally, user response data is collected and sent to the server. This response data indicates how the user reacted to the suggested products. For example, it includes information such as whether the user purchased the product or wrote a review.

[0229] The server analyzes the received response data and provides the results as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, resulting in more accurate suggestions in the future.

[0230] As a concrete example, consider a scenario where a user wears smart glasses and walks around a virtual store, and optimal camping equipment is suggested in real time based on past behavioral data. If the user reviews the suggested products and indicates a purchase intention, that data is sent back to the server and used to improve the accuracy of the model.

[0231] An example of a prompt message is, "Based on the user's past purchase data and browsing history, please display recommended products in the virtual store on the smart glasses in real time."

[0232] In this way, the entire system can improve the user experience by providing highly accurate and timely product recommendations based on user behavior data, and by offering a seamless product selection and purchase experience within a virtual environment.

[0233] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0234] Step 1:

[0235] Collect user behavior data.

[0236] A user is wearing smart glasses and walking around a virtual store. The smart glasses collect real-time information about the user's gaze, gaze duration, and products displayed on the lenses. The input is the user's visual information, and the output is motion data containing this information. This data is temporarily stored within the visual device.

[0237] Step 2:

[0238] Send the collected motion data to the server.

[0239] The visual device collects motion data and sends it to a server. A network connection is used for this transmission. The input is motion data stored in the smart glasses, and the output is motion data that reaches the server. This data transmission is performed using the HTTPS protocol.

[0240] Step 3:

[0241] Save and initialize operation data.

[0242] The server receives operational data, saves it to a database, and performs initial processing. This initial processing includes data cleansing, such as removing noise and filling in missing data. The input is raw operational data, and the output is cleansed and refined operational data.

[0243] Step 4:

[0244] Training an artificial intelligence model

[0245] The server uses pre-processed behavioral data to train an artificial intelligence model. This model learns user preferences based on past purchase and browsing history. The input is cleansed behavioral data, and the output is a trained AI model. Specifically, data analysis algorithms are used to extract important features and improve the model's prediction accuracy.

[0246] Step 5:

[0247] Generate product proposals

[0248] Product suggestions are generated based on a trained artificial intelligence model. The server considers the user's past behavior and current trend information to select the most suitable product for the user. The input consists of a trained AI model and current trend data, and the output is a list of specific product suggestions. The suggestion list includes descriptions and price information for each product.

[0249] Step 6:

[0250] Display product suggestions to users.

[0251] The generated product suggestions are sent from the server to the visual device and displayed to the user in real time. The input is product suggestion data sent from the server, and the output is a list of products displayed on the visual device's screen. The user can review the products and explore those of interest in more detail.

[0252] Step 7:

[0253] Collect response data from users and send it to the server.

[0254] When a user selects or purchases a suggested product, the visual device collects response data and sends it to a server. The input is user interaction information, and the output is response data sent to the server. This data includes purchase history and product reviews.

[0255] Step 8:

[0256] Evaluate response data and update the artificial intelligence model.

[0257] The server analyzes the received response data and feeds the results back into the artificial intelligence model. This allows the model to retrain, resulting in more accurate suggestions for the next time. The input is newly collected response data, and the output is an updated artificial intelligence model. Specifically, the product list is adjusted with an emphasis on positive feedback.

[0258] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0259] This invention is a system that collects user behavioral and emotional data and generates personalized product recommendations based on that data. The system consists of the following main components:

[0260] 1. Collecting user behavior data

[0261] User terminal operation

[0262] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[0263] 2. Collecting user sentiment data

[0264] User terminal operation

[0265] We collect user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, we analyze the emotion from that text. We also use cameras to infer emotions from the user's facial expressions and voice.

[0266] 3. Data transmission and storage

[0267] User terminal operation

[0268] The collected behavioral and emotional data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and emotional state.

[0269] Server operation

[0270] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[0271] 4. Training of the artificial intelligence model

[0272] Server operation

[0273] Using cleansed behavioral and emotional data, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, it learns characteristics that indicate a user prefers "outdoor activities" based on past purchase history, browsing history, and emotional states.

[0274] 5. Generating Product Proposals

[0275] Server operation

[0276] Based on a pre-trained artificial intelligence model, product recommendations are generated for the user. These recommendations take into account the user's behavioral history, emotional data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment might be suggested.

[0277] User terminal operation

[0278] The system receives product data suggested by the server, converts it to a display format, and displays it through the user interface.

[0279] 6. Collecting user responses and feedback

[0280] User actions

[0281] Users can browse suggested products, click on items they're interested in to see more details, add them to their cart, and complete the purchase. After the purchase, they can provide reviews and feedback, generating further data.

[0282] User terminal operation

[0283] This response data is collected, converted to a specified format, and sent to the server. This data includes product ID, rating score, review comments, sentiment status, and more.

[0284] 7. Evaluation of Responses and Update of AI Model

[0285] Server Operations

[0286] The server analyzes the response data received from the user terminal to grasp new preferences, behavior patterns, and emotional patterns. For example, if a user highly evaluates camping supplies of a specific brand and detects a positive emotional state from the review.

[0287] Server Operations

[0288] By retraining the AI model based on the new response data, the accuracy of the model is improved, and the next recommendation is further optimized.

[0289] Specific Example

[0290] For example, when a user who frequently browses and purchases camping supplies is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and emotional data. The server trains the AI model based on these data and generates an optimal list of camping supplies considering that the user enjoys "outdoor activities". This list is sent to the user terminal and displayed through the user interface. If the user purchases the proposed product and posts a positive review, the emotional data is also sent back to the server and used as training data for the AI model, so that subsequent recommendations are further optimized for the user. In this way, the entire system realizes optimal product recommendations while performing self-improvement in real time based on the user's action data and emotional data.

[0291] The following describes the processing flow.

[0292] Step 1:

[0293] User Operations

[0294] A user accesses an e-commerce site and browses products. During this process, information about the viewed products is recorded in real time. For example, a user might view a sports equipment page multiple times and check the details of a specific product.

[0295] Step 2:

[0296] User terminal operation

[0297] The user's terminal converts the collected browsing history into a specified data format. This data includes product ID, browsing time, and page transition order. The converted data is then sent to the server.

[0298] Step 3:

[0299] User terminal operation

[0300] Users write reviews for specific products. Text analysis technology is used to extract sentiment data from these reviews. For example, emotions such as "satisfied" or "dissatisfied" can be identified from keywords and sentence tone used in the reviews.

[0301] Step 4:

[0302] User terminal operation

[0303] The system uses a built-in camera and microphone to collect emotional data from the user's facial expressions and voice while they are browsing products or writing reviews. For example, facial recognition technology is used to determine whether the user is smiling, angry, or otherwise in a negative mood.

[0304] Step 5:

[0305] User terminal operation

[0306] The collected behavioral and emotional data is converted into a specified data format and sent to the server. The converted data includes product ID, viewing time, number of purchases, review content, and emotional state.

[0307] Step 6:

[0308] Server Operations

[0309] The server receives data sent from the user terminal. The received data is stored in the database in real time. During the storage process, data cleansing is performed to correct inconsistent data and supplement missing data.

[0310] Step 7:

[0311] Server Operations

[0312] Using the cleansed operation data and emotion data, train an artificial intelligence model to learn the user's preferences, behavior patterns, and emotion patterns. For example, learn the characteristics to determine a user who is interested in "outdoor activities" from past purchase history, browsing history, and emotional state.

[0313] Step 8:

[0314] Server Operations

[0315] Based on the trained artificial intelligence model, generate product recommendations for the user. Here, recommendations are made considering the user's behavior history, emotion data, and current trend information. For example, as summer approaches, trail running shoes and camping supplies are recommended.

[0316] Step 9:

[0317] User Terminal Operations

[0318] Receive the product recommendation data sent from the server, convert it into a display format, and display it through the user interface. The user can view and check the details of the product recommendations through this interface.

[0319] Step 10:

[0320] User actions

[0321] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. After purchase, they are also required to provide reviews and feedback, documenting their feelings about the product.

[0322] Step 11:

[0323] User terminal operation

[0324] The system collects data on user purchases, reviews, and feedback. This data is converted to a specified format and sent to the server. The data includes product ID, rating score, review comments, and sentiment status.

[0325] Step 12:

[0326] Server operation

[0327] The server analyzes the response data received from the user's terminal. This analysis reveals the user's latest preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[0328] Step 13:

[0329] Server operation

[0330] The AI ​​model is retrained based on the new response data. This retraining improves the accuracy of the AI ​​model, further optimizing future suggestions. New features and patterns are learned, continuously improving the model's performance.

[0331] In this way, the program collects user behavior and emotional data in real time, and uses AI to analyze and make suggestions, thereby improving the user experience.

[0332] (Example 2)

[0333] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0334] Traditional e-commerce systems only collect user behavior data to suggest products, making it difficult to provide personalized recommendations based on user emotions and interests. Furthermore, the accuracy of product recommendations was low, making it difficult to improve user satisfaction. In addition, because user emotional data was not reflected in the system, more appropriate product recommendations could not be achieved.

[0335] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0336] In this invention, the server includes means for collecting user behavior data and emotional data, means for transmitting the collected behavior data and emotional data to the server, and means for storing and initial processing the transmitted behavior data and emotional data. This enables more personalized product recommendations by utilizing both the user's behavior data and emotional data. Specifically, it includes means for collecting emotional data in real time using user emotion recognition technology, means for inferring emotions from the user's facial expressions and voice using a camera, and means for improving the accuracy of product recommendations using the user's emotional data, thereby realizing more accurate product recommendations.

[0337] "Action data" refers to information about user actions such as browsing, purchasing, and providing feedback on e-commerce sites.

[0338] "Emotional data" refers to information about a user's emotional state obtained from facial recognition technology, text analysis, and facial expressions and voice captured using a camera.

[0339] "Means of collection" refers to hardware and software for acquiring user behavioral data and emotional data in real time.

[0340] "Means of transmission" refers to communication means for converting collected behavioral data and emotional data into the required format and sending it to the server.

[0341] "Means for saving and initial processing" refers to the process of saving the behavioral and emotional data sent to the server into a database, verifying the integrity of the data, and cleansing up any inaccurate data.

[0342] "Training methods" refer to methods for building and training artificial intelligence models that learn user preferences, behavioral patterns, and emotional patterns using stored and pre-processed data.

[0343] "Methods for generating product suggestions" refers to methods that utilize trained artificial intelligence models to generate personalized and optimal product suggestions for each user.

[0344] "Means of display" refers to a mechanism for sending generated product suggestions to the user's terminal and presenting them visually through the user interface.

[0345] "Response data" refers to information about actions users take in response to product suggestions (e.g., viewing products, purchasing them, writing reviews).

[0346] "Methods for updating" refer to methods of analyzing collected response data and retraining the artificial intelligence model to improve its accuracy.

[0347] "Emotion recognition technology" is a technology that uses facial recognition and text analysis of users, as well as a camera to analyze facial expressions and voice, in order to infer the user's emotional state.

[0348] This invention is a system that collects user behavior data and emotional data and generates product recommendations that are optimal for each individual user based on that data. This system includes the following main components:

[0349] 1. Collecting user behavior data

[0350] This refers to the user's terminal activity. This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[0351] 2. Collecting user sentiment data

[0352] This involves the operation of the user's terminal. It collects user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, the emotions are analyzed from the text. It also uses the camera to infer emotions from the user's facial expressions and voice. Specifically, emotion recognition software (e.g., Microsoft® Azure® Face API) is used.

[0353] 3. Data transmission and storage

[0354] This refers to the user's terminal activity. The collected activity and sentiment data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and sentiment state.

[0355] This describes the server's operation. The server stores the received data in a database (e.g., Amazon RDS) in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[0356] 4. Training of the artificial intelligence model

[0357] This is the server operation. Using cleansed behavioral and emotional data, it trains an artificial intelligence model that learns user preferences, behavioral patterns, and emotional patterns. For example, it runs a training script written in Python and builds the model using TensorFlow. From past purchase and browsing history, it learns that the user prefers "outdoor activities."

[0358] 5. Generating Product Proposals

[0359] This describes the server's operation. Based on a pre-trained artificial intelligence model, it generates product suggestions for the user. Here, the suggestions are made considering the user's behavioral history, sentiment data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment may be suggested.

[0360] This describes the operation of the user terminal. It receives product data proposed from the server, converts it into a display format, and displays it through the user interface.

[0361] 6. Collecting user responses and feedback

[0362] This describes user behavior. Users browse suggested products, click on items of interest to view details, add them to their cart, and complete the purchase. Further data is generated when users provide reviews and feedback after the purchase.

[0363] This describes the user terminal's actions. It collects this response data, converts it to a specified format, and sends it to the server. This data includes product ID, rating score, review comments, and sentiment status.

[0364] 7. Evaluating responses and updating artificial intelligence models

[0365] This is how the server works. The server analyzes response data received from the user's terminal to understand new preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[0366] This is a server operation. By retraining the AI ​​model based on new response data, the model's accuracy improves, and the next suggestions become even more optimized.

[0367] Specific example

[0368] For example, if a user who frequently browses and purchases camping gear is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and sentiment data. The server uses this data to train an artificial intelligence model and generates an optimal list of camping gear, taking into account the user's enjoyment of "outdoor activities." This list is sent to the user's terminal and displayed through the user interface. If the user purchases the suggested items and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model, further optimizing future suggestions for the user.

[0369] Example of a prompt

[0370] "Please suggest the latest tents to users who frequently browse and purchase camping equipment."

[0371] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0372] Step 1:

[0373] User behavior data collection

[0374] User terminal operation

[0375] We collect this behavioral data when users browse, purchase, and provide feedback on products on e-commerce sites. For example, when a user is browsing a specific sporting item, we record the product ID, the start time of the browsing, and the end time. If the user completes a purchase, we also collect the purchased product ID, the date and time of purchase, and the purchase amount.

[0376] Input: User's behavior on the website

[0377] Output: Operational data such as product ID, viewing time, and purchase data.

[0378] Step 2:

[0379] Collecting user sentiment data

[0380] User terminal operation

[0381] Using facial recognition technology and text analysis, we collect emotional data from users when they write product reviews or use the website. Specifically, we extract emotions from the text of product reviews written by users, and infer emotions from facial expressions and voice using a camera. For this purpose, we use emotion recognition software.

[0382] Input: User review text, camera footage, audio

[0383] Output: Emotional data such as emotion type (positive, negative, etc.) and intensity.

[0384] Step 3:

[0385] Sending and storing data

[0386] User terminal operation

[0387] The collected behavioral and emotional data is converted into JSON format and sent to the server in real time. For example, the dataset might look like this: "Product ID: 1234, Viewing time: 5 minutes, Purchase date and time: 2023-10-12, Review emotion: Positive, Facial emotion: Smile".

[0388] Input: Behavioral data, emotional data

[0389] Output: Data in JSON format

[0390] Server operation

[0391] The server stores the received data in a database and performs a cleansing process as an initial step. This removes inconsistent or missing data to ensure data integrity.

[0392] Input: Data in JSON format

[0393] Output: Cleansed and consistent data

[0394] Step 4:

[0395] Training of artificial intelligence models

[0396] Server operation

[0397] Based on cleansed behavioral and emotional data, an AI model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, a script written in Python is executed, and the model is built using TensorFlow. The training data includes past purchase history, browsing history, and emotional states.

[0398] Input: Data after cleansing

[0399] Output: Trained AI model

[0400] Step 5:

[0401] Product Proposal Generation

[0402] Server operation

[0403] Using a pre-trained artificial intelligence model, the system generates personalized product recommendations for each user. Specifically, it considers the user's past behavioral history, emotional data, and current trend information when making recommendations. For example, "as summer approaches, it might suggest trail running shoes or camping equipment."

[0404] Input: Trained AI model, user behavior data, sentiment data, trend information

[0405] Output: Product Proposal List

[0406] User terminal operation

[0407] The system receives product data suggested by the server and converts it into a format for display in the user interface. This allows the user to visually see product images and detailed information.

[0408] Input: Product Proposal List

[0409] Output: Product list displayed in the user interface

[0410] Step 6:

[0411] Collecting user responses and feedback

[0412] User actions

[0413] Users can view suggested products, check their details, add them to their cart, and complete the purchase. After purchase, they can provide reviews and feedback.

[0414] Input: User actions (browsing, purchasing, reviewing)

[0415] Output: Response data

[0416] User terminal operation

[0417] The collected response data is converted to JSON format and sent to the server. This data includes product ID, rating score, review comments, sentiment status, etc.

[0418] Input: User action data

[0419] Output: Response data in JSON format

[0420] Step 7:

[0421] Response evaluation and AI model updates

[0422] Server operation

[0423] The server analyzes the received response data to understand new user preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[0424] Input: Response data in JSON format

[0425] Output: Updated user preference data, behavioral pattern data, emotional pattern data

[0426] Server operation

[0427] The AI ​​model is retrained based on the new response data to improve its accuracy. This will further optimize future suggestions.

[0428] Input: Updated user preference data, behavioral pattern data, emotional pattern data

[0429] Output: Improved AI model

[0430] (Application Example 2)

[0431] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0432] Traditional e-commerce systems suggested products based on users' purchase and browsing history, but it was difficult to provide optimal product suggestions that took into account user emotions and real-time behavior. Furthermore, the lack of functionality to collect and analyze user emotion data prevented personalized suggestions that reflected user feelings. As a result, improvements in the user experience were limited.

[0433] The specific processing performed 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 user behavior data and emotion data, means for transmitting the collected behavior data and emotion data to the server, and means for storing and initial processing the transmitted behavior data and emotion data. This makes it possible to provide more precise personalized product suggestions based on the user's behavior and emotions.

[0434] "Action data" refers to the user's behavioral history on the e-commerce system, including purchase history, browsing history, and feedback information.

[0435] "Emotional data" refers to information about a user's emotional state obtained through methods such as facial expressions, voice, and text analysis.

[0436] "Personalized product recommendations" refer to product recommendations optimized for each user, generated based on their individual behavioral and emotional data.

[0437] "Facial recognition" refers to a technology that captures a user's facial expressions and analyzes their emotions from those expressions.

[0438] "Text analysis" refers to the technology that analyzes text data entered by users and understands their emotions and intentions from its content.

[0439] "Trend information" refers to data about current market and consumer trends.

[0440] "Response data" refers to feedback information obtained from user actions taken in response to product suggestions (e.g., clicks, purchases, reviews, etc.).

[0441] An "artificial intelligence model" refers to a machine learning model that learns from collected and analyzed data to recognize and predict patterns and trends.

[0442] "Real-time" refers to processing and making suggestions that immediately reflect changes in user behavior and emotions.

[0443] This invention relates to a system that collects user behavior data and emotional data and generates personalized product recommendations based on that data. This system is primarily implemented using the following hardware and software components.

[0444] First, the user's smartphone will have programs installed to collect behavioral and emotional data. Collecting behavioral data includes functions to acquire information (purchase history, browsing history, feedback information, etc.) when the user browses or purchases products on e-commerce sites. Simultaneously, emotional data will be collected using facial recognition and text analysis software utilizing the smartphone's camera. Specifically, the OpenCV library will be used to capture the user's facial expressions and analyze their emotions in real time. Text analysis will use the NLTK library to analyze emotions from text data such as reviews and feedback.

[0445] The collected behavioral and emotional data are converted to a specified data format and sent over the internet to AWS (Amazon Web Services) servers. The servers store the received data in a database such as Amazon RDS and perform data cleansing using Python scripts. After data cleansing, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. Machine learning algorithms such as Scikit-learn's Random Forest Classifier are used to train this model.

[0446] Next, a trained artificial intelligence model is used to generate personalized product recommendations for the user. This process takes into account the user's past browsing and purchase history, emotional state, and market trend information. The generated product recommendations are sent to the user's smartphone in real time and displayed to the user through an appropriate interface.

[0447] When a user takes action regarding a suggested product (e.g., clicking on a product, purchasing it, or posting a review), detailed response data is collected again and sent to the server. This response data is used to understand new preferences, behavioral patterns, and emotional patterns, and the AI ​​model is further updated and improved.

[0448] As a concrete example, suppose a user is browsing camping equipment and looking for a new tent. In this case, the system collects the user's browsing history, purchase history, reviews, and sentiment data. On the server side, an artificial intelligence model is trained using this data to generate a list of optimal camping equipment for the user, who is determined to prefer "outdoor activities." This list is sent to the smartphone and displayed through the user interface. If the user purchases the suggested products and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model.

[0449] Here are some concrete examples of prompt statements:

[0450] "When users write reviews about camping equipment, use review comments and facial expression data to determine their emotional state."

[0451] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0452] Step 1:

[0453] When a user browses an e-commerce site using their smartphone, the smartphone device collects operational data. This data includes the ID of the product viewed, the viewing time, and the number of clicks. This data is collected and stored in a local database.

[0454] Step 2:

[0455] When users write product reviews, their smartphone cameras and text input are used to collect emotional data. The input includes user facial expressions and review comments. The OpenCV library is used to analyze the facial expressions and infer the user's emotional state. Simultaneously, the NLTK library is used to analyze the review comments and recognize the emotions expressed in the text. This emotional data is also stored in a local database.

[0456] Step 3:

[0457] The smartphone device converts collected behavioral and emotional data into a specified data format and sends it to the server via the internet. The input for this process is all behavioral and emotional data stored locally. The output is the transmission process of this data. The data transmission is performed securely using the HTTPS protocol.

[0458] Step 4:

[0459] The server stores the received behavioral and sentimental data in a database such as Amazon RDS and performs initial processing using a Python script. The input consists of behavioral and sentimental data sent from the user's terminal. Initial processing includes data cleansing and formatting. The output is the cleaned data.

[0460] Step 5:

[0461] The server uses the cleansed data to train an artificial intelligence model. The inputs include cleansed behavioral data and sentiment data. A machine learning model is built and trained using tools such as Scikit-learn's Random Forest. The output is the trained AI model.

[0462] Step 6:

[0463] Based on a trained AI model, the server generates personalized product recommendations. Inputs include the trained AI model and newly submitted user behavior and sentiment data. Market trend information is also incorporated. The AI ​​model generates an optimized product list. The output is a product recommendation list.

[0464] Step 7:

[0465] The server sends the generated product suggestion list to the user's smartphone. The input is the product suggestion list generated by the AI ​​model. The output is the product suggestion displayed on the user's smartphone.

[0466] Step 8:

[0467] Action and sentiment data are collected again when the user purchases or clicks on one of the suggested products. The input is the user's new behavior and sentiment data. This data is again stored in the local database and prepared to be sent to the server.

[0468] Step 9:

[0469] The server receives new user behavior and emotion data and updates the AI ​​model. The input is the latest behavior and emotion data from the user. The AI ​​model is retrained using this new data, enabling it to make more accurate product recommendations. The output is the updated AI model.

[0470] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating 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.

[0471] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0472] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0473] [Second Embodiment]

[0474] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0475] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0476] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0477] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0478] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0479] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0480] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0481] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0482] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0483] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0484] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0485] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. 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".

[0486] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components.

[0487] 1. Collecting user behavior data

[0488] User terminal operation

[0489] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[0490] 2. Data transmission and storage

[0491] User terminal operation

[0492] The collected behavioral data is converted into a format and sent to the server. For example, it may include information such as the frequency of access to a specific product category and the frequency of purchases.

[0493] Server operation

[0494] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[0495] 3. Training of artificial intelligence models

[0496] Server operation

[0497] Using cleansed data, an artificial intelligence model is trained to learn user preferences and behavioral patterns. For example, past purchase and browsing history is used to learn features that indicate a user's interest in outdoor activities.

[0498] 4. Generating Product Proposals

[0499] Server operation

[0500] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior and current trends. For example, as summer approaches, recommendations for trail running shoes and camping gear will be made.

[0501] User terminal operation

[0502] The user terminal receives the suggestion data sent from the server and displays it through the user interface.

[0503] 5. Collecting user responses and feedback

[0504] User actions

[0505] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[0506] User terminal operation

[0507] This response data is collected and sent to the server.

[0508] 6. Evaluating responses and updating artificial intelligence models

[0509] Server operation

[0510] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, the characteristics of products that users have left particularly positive reviews on are taken into consideration and reflected in subsequent recommendations.

[0511] Specific example

[0512] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed to them. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations. In this way, the entire system is designed to continuously improve itself to enhance the user experience.

[0513] The following describes the processing flow.

[0514] Step 1:

[0515] User actions

[0516] When a user accesses an e-commerce site and browses products, information about the viewed products is recorded in real time. This includes actions such as the user navigating to the product details page, checking reviews, and adding the product to their cart. Furthermore, if the user completes a purchase, that purchase information is also recorded.

[0517] Step 2:

[0518] User terminal operation

[0519] The user's terminal converts collected browsing history, purchase history, and review information into a specified data format. This data includes details such as product ID, viewing time, number of purchases, and review content. The converted data is then sent to the server.

[0520] Step 3:

[0521] Server operation

[0522] The server receives data sent from the user's terminal. The received data is saved to the database in real time. During the saving process, data cleansing is performed. This cleansing process includes supplementing missing data and correcting inconsistent data.

[0523] Step 4:

[0524] Server operation

[0525] The AI ​​model is trained using the cleansed data. This AI model extracts features from user behavior data and learns user preferences and patterns. For example, it learns that a user likes "outdoor activities" based on their past purchase and browsing history.

[0526] Step 5:

[0527] Server operation

[0528] Using a trained AI model, product recommendations are generated for the user. These recommendations combine the user's past behavior with current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[0529] Step 6:

[0530] User terminal operation

[0531] The system receives product suggestion data sent from the server and converts it into a display format. The converted product suggestion list is then displayed to the user through an easy-to-view interface. The user can then browse the suggested products and decide whether or not to purchase them.

[0532] Step 7:

[0533] User actions

[0534] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. Further data is generated when users provide reviews and feedback after purchase.

[0535] Step 8:

[0536] User terminal operation

[0537] The system collects data on user purchases, reviews, and feedback. This data is then converted back into a specified format and sent to the server. This data includes product IDs, rating scores, and review comments.

[0538] Step 9:

[0539] Server operation

[0540] The server analyzes the response data received from the user's terminal. This analysis helps to understand the user's latest preferences and behavioral patterns. For example, if a user gives a high rating to a particular brand of camping equipment, that information will be used as a basis for the analysis.

[0541] Step 10:

[0542] Server operation

[0543] The AI ​​model is retrained based on the new response data. This updates the AI ​​model to reflect the latest data, enabling more accurate product recommendations in the future. New features and patterns are learned, continuously improving the model's performance.

[0544] In this way, the program aims to improve the user experience by collecting user behavior data in real time and using AI to analyze and make suggestions.

[0545] (Example 1)

[0546] Next, we will describe Example 1. 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".

[0547] Traditional e-commerce systems have struggled to provide optimal product recommendations based on individual user preferences and behavior. In particular, the lack of mechanisms to collect and rapidly analyze user behavior data in real time for product recommendations led to a decline in the quality of the user experience. Furthermore, insufficient data cleansing processes to ensure the accuracy of collected data meant that improving the accuracy of artificial intelligence models remained a challenge.

[0548] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0549] In this invention, the server includes means for storing and cleaning collected behavioral data, means for training an artificial intelligence model that learns using the stored and cleansed behavioral data, and means for generating product suggestions based on the trained artificial intelligence model. This makes it possible to provide more accurate product suggestions in real time based on the behavioral data of individual users.

[0550] "User activity data" refers to information about all actions and behaviors that a user performs on a website or application. This includes browsing history, purchase history, feedback, and access frequency.

[0551] A "server" refers to a computer system that receives, stores, and processes data sent from user terminals, trains artificial intelligence models, and generates product suggestions.

[0552] "Means of collection" refers to methods and devices for capturing and recording user behavior data. This includes data capture by browser scripts and mobile applications.

[0553] "Means of transmission" refers to the means of transferring collected behavioral data from the user's terminal to the server. This includes HTTP requests and API calls.

[0554] "Means for saving and cleansing data" refers to means of saving the transmitted data to a database and performing data cleansing such as imputing missing values ​​and checking data types.

[0555] An "artificial intelligence model" refers to an algorithm or system that learns user behavior patterns and preferences through machine learning or deep learning based on data.

[0556] "Training methods" refer to the methods and devices used to train artificial intelligence models using cleansed data. This includes machine learning frameworks (e.g., TensorFlow and PyTorch).

[0557] "Methods for generating product suggestions" refers to methods that use a trained artificial intelligence model to select the most suitable products for the user and create suggestions.

[0558] "Means of displaying through a user interface" refers to methods and devices for visually presenting generated product suggestions to the user. This includes user interfaces for web pages and mobile applications.

[0559] "Response data" refers to data such as user feedback, purchase history, and ratings regarding suggested products.

[0560] "Means of analysis and updating" refers to methods of analyzing collected response data and retraining an artificial intelligence model based on the results to improve its accuracy.

[0561] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components and processing steps.

[0562] User behavior data collection

[0563] User actions

[0564] User behavioral data is collected when users browse, purchase, and provide feedback on products on e-commerce sites. If a user frequently browses a specific product category (e.g., sporting goods) and purchases related products, that data is collected in real time. Browser scripts and mobile applications are used for this purpose.

[0565] Sending and storing data

[0566] User terminal operation

[0567] The collected behavioral data is converted to JSON format and sent to the server via an HTTP request. For example, it may include logs of when a user frequently viewed a particular product category and information about products they purchased.

[0568] Server operation

[0569] The server stores the received data in a database such as MongoDB or MySQL. The stored data is then cleansed to maintain data accuracy, including imputation of missing values ​​and data type checks.

[0570] Training of artificial intelligence models

[0571] Server operation

[0572] Based on the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow and PyTorch. For example, features are extracted from a user's past purchase and browsing history to determine if the user is interested in outdoor activities.

[0573] Product Proposal Generation

[0574] Server operation

[0575] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior data and current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[0576] User terminal operation

[0577] The user terminal receives the suggestion data sent from the server and displays it through the user interface. For example, a product list might be displayed in the format, "Here are some camping items we recommend for you."

[0578] Collecting user responses and feedback

[0579] User actions

[0580] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[0581] User terminal operation

[0582] This response data is collected and sent to the server. For example, it may include data such as "I purchased this product" or review information such as "I gave it a 5-star rating."

[0583] Response evaluation and AI model updates

[0584] Server operation

[0585] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, it learns the characteristics of products that users have left particularly positive reviews on and incorporates this into subsequent recommendations.

[0586] Specific example

[0587] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed through the user interface. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations.

[0588] Examples of prompts to input into a generative AI model

[0589] "Generate suggestions recommending new camping equipment, scheduled for release next month, to users who have previously purchased outdoor gear."

[0590] "Create a list of optimal product suggestions based on the sports equipment that users frequently view."

[0591] Thus, the present invention is a system that uses user behavior data to make optimal product recommendations and improve the user experience.

[0592] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0593] Step 1:

[0594] User behavior data collection

[0595] User actions

[0596] Users browse e-commerce sites, purchase products, and provide feedback. This behavioral data is captured by each user's device (browser or mobile application). For example, if a user browses products in the "camping equipment" category for more than 10 minutes and adds a specific product to their cart, that information will be collected.

[0597] input

[0598] User browsing history, purchase history, and feedback information

[0599] output

[0600] Local cache of collected behavioral data

[0601] Step 2:

[0602] Sending data

[0603] User terminal operation

[0604] Behavioral data stored in the local cache is converted to JSON format and sent to the server using an HTTP request. For example, JSON data containing access frequency and purchase information for a specific category is generated and sent to the server.

[0605] input

[0606] Local cache behavioral data

[0607] output

[0608] Action data in JSON format sent to the server

[0609] Step 3:

[0610] Data storage and cleansing process

[0611] Server operation

[0612] The server saves the received JSON data to a database (e.g., MongoDB or MySQL). Afterward, a cleansing process is performed, including data imputation and data type checks. For example, if there are missing values ​​in the purchase history, they will be appropriately imputed.

[0613] input

[0614] Behavioral data in JSON format that was sent.

[0615] output

[0616] Cleansed database behavioral data

[0617] Step 4:

[0618] Training of artificial intelligence models

[0619] Server operation

[0620] Using the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow or PyTorch. Here, behavioral patterns and preferences are learned based on the user's past purchase and browsing history. For example, the pattern of a user frequently purchasing "outdoor equipment" is learned.

[0621] input

[0622] Cleansed behavioral data

[0623] output

[0624] Trained artificial intelligence model

[0625] Step 5:

[0626] Product Proposal Generation

[0627] Server operation

[0628] Using a trained model, the system generates optimal product recommendations based on current user data (such as recent browsing history and purchase information). For example, it might suggest new camping equipment to a user interested in "outdoor goods."

[0629] input

[0630] Current user data, trained artificial intelligence model

[0631] output

[0632] Generated product suggestion data (e.g., JSON format)

[0633] Step 6:

[0634] Product suggestion display

[0635] User terminal operation

[0636] The system analyzes product suggestion data received from the server and displays it through the user interface. For example, it can display a list of "recommended camping gear" on a webpage or mobile app screen.

[0637] input

[0638] Product proposal data sent from the server

[0639] output

[0640] Product suggestions displayed to the user

[0641] Step 7:

[0642] Collecting user responses

[0643] User actions

[0644] Users purchase suggested products, write reviews, and provide feedback. These actions are captured as response data.

[0645] input

[0646] User actions (purchase, review, feedback)

[0647] output

[0648] Local cache of collected response data

[0649] Step 8:

[0650] Sending and analyzing response data

[0651] User terminal operation

[0652] The collected response data is sent to the server. The data is then converted back to JSON format and sent via an HTTP request.

[0653] Server operation

[0654] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, improving the accuracy of future suggestions. For example, the characteristics of products that received high ratings can be reflected in future suggestions.

[0655] input

[0656] Response data sent from the user terminal

[0657] output

[0658] AI model updated based on feedback

[0659] (Application Example 1)

[0660] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0661] Traditional e-commerce systems collect user behavior data to suggest products, but they have shortcomings in the accuracy and timeliness of these suggestions. Furthermore, it was difficult to evaluate in real time how beneficial the suggested products were to the user and incorporate that feedback into future suggestions. Additionally, the lack of smooth integration between product suggestions and the purchase process within the virtual environment resulted in a limited user experience.

[0662] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0663] In this invention, the server includes means for collecting user behavior data, means for transmitting the collected behavior data to the server, means for storing and initial processing the transmitted behavior data, means for training an artificial intelligence model that learns using the stored and initial processed behavior data, means for generating product suggestions based on the trained artificial intelligence model, means for displaying the generated product suggestions to the user, means for collecting and transmitting response data from the user to the server, means for evaluating the collected response data and updating the artificial intelligence model, and means for displaying product suggestions via a visual device used by the user in a virtual environment, enabling the selection and purchase of suggested products. This enables highly accurate and timely product suggestions based on user behavior data, and provides a seamless product selection and purchase experience in a virtual environment.

[0664] "Action data" refers to data about a user's actions and choices within an online or virtual environment, such as browsing history, purchase history, and feedback information.

[0665] A "server" refers to a computer system that receives, stores, and processes operational data over a network, and is used for training artificial intelligence models and generating product suggestions.

[0666] An "artificial intelligence model" refers to software that includes algorithms that analyze user preferences and behavior based on trained data and provide optimal product recommendations.

[0667] "Product recommendations" refer to information that indicates products or services recommended to the user, based on collected behavioral data and analysis results from artificial intelligence models.

[0668] "Visual devices" refer to devices that users wear to display information in a virtual environment, such as smart glasses and head-mounted displays.

[0669] "Response data" refers to information that shows users' reactions and feedback to suggested products, such as whether they purchased the product or wrote an article or review.

[0670] "Training" refers to the process of using collected and initially processed behavioral data to train an artificial intelligence model, thereby improving its prediction and suggestion accuracy.

[0671] A "virtual environment" refers to a virtual commercial space or store that a user experiences through a visual device using virtual reality technology.

[0672] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system improves the user experience through the following steps.

[0673] The system first collects user behavior data. This is done using visual devices such as smart glasses or head-mounted displays. These devices collect the user's gaze and gaze duration in real time, recording what the user is looking at.

[0674] The collected motion data is sent from the visual device to the server. The server stores the received data in a database and performs initial processing. This initial processing includes data cleansing, for example, removing noise from the gaze data and extracting accurate motion patterns.

[0675] Next, the server uses the cleansed data to train an artificial intelligence model. This training utilizes past purchase and browsing history. The trained AI model learns the user's preferences and behavioral patterns and generates optimal product recommendations.

[0676] Once product suggestions are generated, the server sends them to the user's visual device for display. The user can then review, select, and purchase the suggested products through their visual device. This allows the user to have a seamless product selection and purchase experience within the virtual environment.

[0677] Additionally, user response data is collected and sent to the server. This response data indicates how the user reacted to the suggested products. For example, it includes information such as whether the user purchased the product or wrote a review.

[0678] The server analyzes the received response data and provides the results as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, resulting in more accurate suggestions in the future.

[0679] As a concrete example, consider a scenario where a user wears smart glasses and walks around a virtual store, and optimal camping equipment is suggested in real time based on past behavioral data. If the user reviews the suggested products and indicates a purchase intention, that data is sent back to the server and used to improve the accuracy of the model.

[0680] An example of a prompt message is, "Based on the user's past purchase data and browsing history, please display recommended products in the virtual store on the smart glasses in real time."

[0681] In this way, the entire system can improve the user experience by providing highly accurate and timely product recommendations based on user behavior data, and by offering a seamless product selection and purchase experience within a virtual environment.

[0682] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0683] Step 1:

[0684] Collect user behavior data.

[0685] A user is wearing smart glasses and walking around a virtual store. The smart glasses collect real-time information about the user's gaze, gaze duration, and products displayed on the lenses. The input is the user's visual information, and the output is motion data containing this information. This data is temporarily stored within the visual device.

[0686] Step 2:

[0687] Send the collected motion data to the server.

[0688] The visual device collects motion data and sends it to a server. A network connection is used for this transmission. The input is motion data stored in the smart glasses, and the output is motion data that reaches the server. This data transmission is performed using the HTTPS protocol.

[0689] Step 3:

[0690] Save and initialize operation data.

[0691] The server receives operational data, saves it to a database, and performs initial processing. This initial processing includes data cleansing, such as removing noise and filling in missing data. The input is unprocessed operational data, and the output is cleansed and refined operational data.

[0692] Step 4:

[0693] Training an artificial intelligence model

[0694] The server uses pre-processed behavioral data to train an artificial intelligence model. This model learns user preferences based on past purchase and browsing history. The input is cleansed behavioral data, and the output is a trained AI model. Specifically, data analysis algorithms are used to extract important features and improve the model's prediction accuracy.

[0695] Step 5:

[0696] Generate product proposals

[0697] Product suggestions are generated based on a trained artificial intelligence model. The server considers the user's past behavior and current trend information to select the most suitable product for the user. The input consists of a trained AI model and current trend data, and the output is a list of specific product suggestions. The suggestion list includes descriptions and price information for each product.

[0698] Step 6:

[0699] Display product suggestions to users.

[0700] The generated product suggestions are sent from the server to the visual device and displayed to the user in real time. The input is product suggestion data sent from the server, and the output is a list of products displayed on the visual device's screen. The user can review the products and explore those of interest in more detail.

[0701] Step 7:

[0702] Collect response data from users and send it to the server.

[0703] When a user selects or purchases a suggested product, the visual device collects response data and sends it to a server. The input is user interaction information, and the output is response data sent to the server. This data includes purchase history and product reviews.

[0704] Step 8:

[0705] Evaluate response data and update the artificial intelligence model.

[0706] The server analyzes the received response data and feeds the results back into the artificial intelligence model. This allows the model to retrain, resulting in more accurate suggestions for the next time. The input is newly collected response data, and the output is an updated artificial intelligence model. Specifically, the product list is adjusted with an emphasis on positive feedback.

[0707] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0708] This invention is a system that collects user behavioral and emotional data and generates personalized product recommendations based on that data. The system consists of the following main components:

[0709] 1. Collecting user behavior data

[0710] User terminal operation

[0711] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[0712] 2. Collecting user sentiment data

[0713] User terminal operation

[0714] We collect user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, we analyze the emotion from that text. We also use cameras to infer emotions from the user's facial expressions and voice.

[0715] 3. Data transmission and storage

[0716] User terminal operation

[0717] The collected behavioral and emotional data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and emotional state.

[0718] Server operation

[0719] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[0720] 4. Training of the artificial intelligence model

[0721] Server operation

[0722] Using cleansed behavioral and emotional data, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, it learns characteristics that indicate a user prefers "outdoor activities" based on past purchase history, browsing history, and emotional states.

[0723] 5. Generating Product Proposals

[0724] Server operation

[0725] Based on a pre-trained artificial intelligence model, product recommendations are generated for the user. These recommendations take into account the user's behavioral history, emotional data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment might be suggested.

[0726] User terminal operation

[0727] The system receives product data suggested by the server, converts it to a display format, and displays it through the user interface.

[0728] 6. Collecting user responses and feedback

[0729] User actions

[0730] Users can browse suggested products, click on items they're interested in to see more details, add them to their cart, and complete the purchase. After the purchase, they can provide reviews and feedback, generating further data.

[0731] User terminal operation

[0732] This response data is collected, converted to a specified format, and sent to the server. This data includes product ID, rating score, review comments, sentiment status, and more.

[0733] 7. Evaluating responses and updating artificial intelligence models

[0734] Server operation

[0735] The server analyzes response data received from user terminals to understand new preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[0736] Server operation

[0737] By retraining the AI ​​model based on new response data, the model's accuracy improves, and the next suggestions become even more optimized.

[0738] Specific example

[0739] For example, if a user who frequently browses and purchases camping equipment is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and sentiment data. The server trains an artificial intelligence model based on this data and generates an optimal list of camping equipment, taking into account the user's enjoyment of "outdoor activities." This list is sent to the user's terminal and displayed through the user interface. If the user purchases the suggested products and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model, further optimizing future suggestions for the user. In this way, the entire system achieves optimal product recommendations while continuously improving in real time based on the user's behavioral and sentiment data.

[0740] The following describes the processing flow.

[0741] Step 1:

[0742] User actions

[0743] A user accesses an e-commerce site and browses products. During this process, information about the viewed products is recorded in real time. For example, a user might view a sports equipment page multiple times and check the details of a specific product.

[0744] Step 2:

[0745] User terminal operation

[0746] The user's terminal converts the collected browsing history into a specified data format. This data includes product ID, browsing time, and page transition order. The converted data is then sent to the server.

[0747] Step 3:

[0748] User terminal operation

[0749] Users write reviews for specific products. Text analysis technology is used to extract sentiment data from these reviews. For example, emotions such as "satisfied" or "dissatisfied" are identified based on keywords and sentence tone used in the reviews.

[0750] Step 4:

[0751] User terminal operation

[0752] The system uses a built-in camera and microphone to collect emotional data from the user's facial expressions and voice while they are browsing products or writing reviews. For example, facial recognition technology is used to determine whether the user is smiling, angry, or otherwise in a negative mood.

[0753] Step 5:

[0754] User terminal operation

[0755] The collected behavioral and emotional data is converted into a specified data format and sent to the server. The converted data includes product ID, viewing time, number of purchases, review content, and emotional state.

[0756] Step 6:

[0757] Server operation

[0758] The server receives data sent from the user's terminal. The received data is saved to the database in real time. During the saving process, data cleansing is performed, correcting inconsistent data and supplementing missing data.

[0759] Step 7:

[0760] Server operation

[0761] Using cleansed behavioral and emotional data, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, it learns characteristics that identify a user as being interested in "outdoor activities" based on past purchase history, browsing history, and emotional state.

[0762] Step 8:

[0763] Server operation

[0764] Based on a pre-trained artificial intelligence model, product recommendations are generated for the user. These recommendations take into account the user's behavioral history, emotional data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment might be suggested.

[0765] Step 9:

[0766] User terminal operation

[0767] The system receives product suggestion data sent from the server, converts it into a display format, and displays it through the user interface. Users can review the product suggestions and view details through this interface.

[0768] Step 10:

[0769] User actions

[0770] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. After purchase, they are also required to provide reviews and feedback, documenting their feelings about the product.

[0771] Step 11:

[0772] User terminal operation

[0773] The system collects data on user purchases, reviews, and feedback. This data is converted to a specified format and sent to the server. The data includes product ID, rating score, review comments, and sentiment status.

[0774] Step 12:

[0775] Server operation

[0776] The server analyzes the response data received from the user's terminal. This analysis reveals the user's latest preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[0777] Step 13:

[0778] Server operation

[0779] The AI ​​model is retrained based on the new response data. This retraining improves the accuracy of the AI ​​model, further optimizing future suggestions. New features and patterns are learned, continuously improving the model's performance.

[0780] In this way, the program collects user behavior and emotional data in real time, and uses AI to analyze and make suggestions, thereby improving the user experience.

[0781] (Example 2)

[0782] Next, we will describe Example 2. 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".

[0783] Traditional e-commerce systems only collect user behavior data to suggest products, making it difficult to provide personalized recommendations based on user emotions and interests. Furthermore, the accuracy of product recommendations was low, making it difficult to improve user satisfaction. In addition, because user emotional data was not reflected in the system, more appropriate product recommendations could not be achieved.

[0784] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0785] In this invention, the server includes means for collecting user behavior data and emotional data, means for transmitting the collected behavior data and emotional data to the server, and means for storing and initial processing the transmitted behavior data and emotional data. This enables more personalized product recommendations by utilizing both the user's behavior data and emotional data. Specifically, it includes means for collecting emotional data in real time using user emotion recognition technology, means for inferring emotions from the user's facial expressions and voice using a camera, and means for improving the accuracy of product recommendations using the user's emotional data, thereby realizing more accurate product recommendations.

[0786] "Action data" refers to information about user actions such as browsing, purchasing, and providing feedback on e-commerce sites.

[0787] "Emotional data" refers to information about a user's emotional state obtained from facial recognition technology, text analysis, and facial expressions and voice captured using a camera.

[0788] "Means of collection" refers to hardware and software for acquiring user behavioral data and emotional data in real time.

[0789] "Means of transmission" refers to communication means for converting collected behavioral data and emotional data into the required format and sending it to the server.

[0790] "Means for saving and initial processing" refers to the process of saving the behavioral and emotional data sent to the server into a database, verifying the integrity of the data, and cleansing up any inaccurate data.

[0791] "Training methods" refer to methods for building and training artificial intelligence models that learn user preferences, behavioral patterns, and emotional patterns using stored and pre-processed data.

[0792] "Methods for generating product suggestions" refers to methods that utilize trained artificial intelligence models to generate personalized and optimal product suggestions for each user.

[0793] "Means of display" refers to a mechanism for sending generated product suggestions to the user's terminal and presenting them visually through the user interface.

[0794] "Response data" refers to information about actions users take in response to product suggestions (e.g., viewing products, purchasing them, writing reviews).

[0795] "Methods for updating" refer to methods of analyzing collected response data and retraining the artificial intelligence model to improve its accuracy.

[0796] "Emotion recognition technology" is a technology that uses facial recognition and text analysis of users, as well as a camera to analyze facial expressions and voice, in order to infer the user's emotional state.

[0797] This invention is a system that collects user behavior data and emotional data and generates product recommendations that are optimal for each individual user based on that data. This system includes the following main components:

[0798] 1. Collecting user behavior data

[0799] This refers to the user's terminal activity. This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[0800] 2. Collecting user sentiment data

[0801] This involves the operation of the user's terminal. It collects user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, the emotions are analyzed from the text. It also uses the camera to infer emotions from the user's facial expressions and voice. Specifically, emotion recognition software (e.g., Microsoft Azure Face API) is used.

[0802] 3. Data transmission and storage

[0803] This refers to the user's terminal activity. The collected activity and sentiment data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and sentiment state.

[0804] This describes the server's operation. The server stores the received data in a database (e.g., Amazon RDS) in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[0805] 4. Training of the artificial intelligence model

[0806] This is the server operation. Using cleansed behavioral and emotional data, it trains an artificial intelligence model that learns user preferences, behavioral patterns, and emotional patterns. For example, it runs a training script written in Python and builds the model using TensorFlow. From past purchase and browsing history, it learns that the user prefers "outdoor activities."

[0807] 5. Generating Product Proposals

[0808] This describes the server's operation. Based on a pre-trained artificial intelligence model, it generates product suggestions for the user. Here, the suggestions are made considering the user's behavioral history, sentiment data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment may be suggested.

[0809] This describes the operation of the user terminal. It receives product data proposed from the server, converts it into a display format, and displays it through the user interface.

[0810] 6. Collecting user responses and feedback

[0811] This describes user behavior. Users browse suggested products, click on items of interest to view details, add them to their cart, and complete the purchase. Further data is generated when users provide reviews and feedback after the purchase.

[0812] This describes the user terminal's actions. It collects this response data, converts it to a specified format, and sends it to the server. This data includes product ID, rating score, review comments, and sentiment status.

[0813] 7. Evaluating responses and updating artificial intelligence models

[0814] This is how the server works. The server analyzes response data received from the user's terminal to understand new preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[0815] This is a server operation. By retraining the AI ​​model based on new response data, the model's accuracy improves, and the next suggestions become even more optimized.

[0816] Specific example

[0817] For example, if a user who frequently browses and purchases camping gear is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and sentiment data. The server uses this data to train an artificial intelligence model and generates an optimal list of camping gear, taking into account the user's enjoyment of "outdoor activities." This list is sent to the user's terminal and displayed through the user interface. If the user purchases the suggested items and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model, further optimizing future suggestions for the user.

[0818] Example of a prompt

[0819] "Please suggest the latest tents to users who frequently browse and purchase camping equipment."

[0820] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0821] Step 1:

[0822] User behavior data collection

[0823] User terminal operation

[0824] We collect this behavioral data when users browse, purchase, and provide feedback on products on e-commerce sites. For example, when a user is browsing a specific sporting item, we record the product ID, the start time of the browsing, and the end time. If the user completes a purchase, we also collect the purchased product ID, the date and time of purchase, and the purchase amount.

[0825] Input: User's behavior on the website

[0826] Output: Operational data such as product ID, viewing time, and purchase data.

[0827] Step 2:

[0828] Collecting user sentiment data

[0829] User terminal operation

[0830] Using facial recognition technology and text analysis, we collect emotional data from users when they write product reviews or use the website. Specifically, we extract emotions from the text of product reviews written by users, and infer emotions from facial expressions and voice using a camera. For this purpose, we use emotion recognition software.

[0831] Input: User review text, camera footage, audio

[0832] Output: Emotional data such as type of emotion (positive, negative, etc.) and intensity.

[0833] Step 3:

[0834] Sending and storing data

[0835] User terminal operation

[0836] The collected behavioral and emotional data is converted into JSON format and sent to the server in real time. For example, the dataset might look like this: "Product ID: 1234, Viewing time: 5 minutes, Purchase date and time: 2023-10-12, Review emotion: Positive, Facial emotion: Smile".

[0837] Input: Behavioral data, emotional data

[0838] Output: Data in JSON format

[0839] Server operation

[0840] The server stores the received data in a database and performs a cleansing process as an initial step. This removes inconsistent or missing data to ensure data integrity.

[0841] Input: Data in JSON format

[0842] Output: Cleansed and consistent data

[0843] Step 4:

[0844] Training of artificial intelligence models

[0845] Server operation

[0846] Based on cleansed behavioral and emotional data, an AI model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, a script written in Python is executed, and the model is built using TensorFlow. The training data includes past purchase history, browsing history, and emotional states.

[0847] Input: Data after cleansing

[0848] Output: Trained AI model

[0849] Step 5:

[0850] Product Proposal Generation

[0851] Server operation

[0852] Using a pre-trained artificial intelligence model, the system generates personalized product recommendations for each user. Specifically, it considers the user's past behavioral history, emotional data, and current trend information when making recommendations. For example, "as summer approaches, it might suggest trail running shoes or camping equipment."

[0853] Input: Trained AI model, user behavior data, sentiment data, trend information

[0854] Output: Product Proposal List

[0855] User terminal operation

[0856] The system receives product data suggested by the server and converts it into a format for display in the user interface. This allows the user to visually see product images and detailed information.

[0857] Input: Product Proposal List

[0858] Output: Product list displayed in the user interface

[0859] Step 6:

[0860] Collecting user responses and feedback

[0861] User actions

[0862] Users can view suggested products, check their details, add them to their cart, and complete the purchase. After purchase, they can provide reviews and feedback.

[0863] Input: User actions (browsing, purchasing, reviewing)

[0864] Output: Response data

[0865] User terminal operation

[0866] The collected response data is converted to JSON format and sent to the server. This data includes product ID, rating score, review comments, sentiment status, etc.

[0867] Input: User action data

[0868] Output: Response data in JSON format

[0869] Step 7:

[0870] Response evaluation and AI model updates

[0871] Server operation

[0872] The server analyzes the received response data to understand new user preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[0873] Input: Response data in JSON format

[0874] Output: Updated user preference data, behavioral pattern data, emotional pattern data

[0875] Server operation

[0876] The AI ​​model is retrained based on the new response data to improve its accuracy. This will further optimize future suggestions.

[0877] Input: Updated user preference data, behavioral pattern data, emotional pattern data

[0878] Output: Improved AI model

[0879] (Application Example 2)

[0880] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0881] Traditional e-commerce systems suggested products based on users' purchase and browsing history, but it was difficult to provide optimal product suggestions that took into account user emotions and real-time behavior. Furthermore, the lack of functionality to collect and analyze user emotion data prevented personalized suggestions that reflected user feelings. As a result, improvements in the user experience were limited.

[0882] The specific processing performed 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 user behavior data and emotion data, means for transmitting the collected behavior data and emotion data to the server, and means for storing and initial processing the transmitted behavior data and emotion data. This makes it possible to provide more precise personalized product suggestions based on the user's behavior and emotions.

[0883] "Action data" refers to the user's behavioral history on the e-commerce system, including purchase history, browsing history, and feedback information.

[0884] "Emotional data" refers to information about a user's emotional state obtained through methods such as facial expressions, voice, and text analysis.

[0885] "Personalized product recommendations" refer to product recommendations optimized for each user, generated based on their individual behavioral and emotional data.

[0886] "Facial recognition" refers to a technology that captures a user's facial expressions and analyzes their emotions from those expressions.

[0887] "Text analysis" refers to the technology that analyzes text data entered by users and understands their emotions and intentions from its content.

[0888] "Trend information" refers to data about current market and consumer trends.

[0889] "Response data" refers to feedback information obtained from user actions taken in response to product suggestions (e.g., clicks, purchases, reviews, etc.).

[0890] An "artificial intelligence model" refers to a machine learning model that learns from collected and analyzed data to recognize and predict patterns and trends.

[0891] "Real-time" refers to processing and making suggestions that immediately reflect changes in user behavior and emotions.

[0892] This invention relates to a system that collects user behavior data and emotional data and generates personalized product recommendations based on that data. This system is primarily implemented using the following hardware and software components.

[0893] First, the user's smartphone will have programs installed to collect behavioral and emotional data. Collecting behavioral data includes functions to acquire information (purchase history, browsing history, feedback information, etc.) when the user browses or purchases products on e-commerce sites. Simultaneously, emotional data will be collected using facial recognition and text analysis software utilizing the smartphone's camera. Specifically, the OpenCV library will be used to capture the user's facial expressions and analyze their emotions in real time. Text analysis will use the NLTK library to analyze emotions from text data such as reviews and feedback.

[0894] The collected behavioral and emotional data are converted to a specified data format and sent to an AWS (Amazon Web Services) server via the internet. The server stores the received data in a database such as Amazon RDS and performs data cleansing using a Python script. After the data is cleansed, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. Machine learning algorithms such as Scikit-learn's Random Forest Classifier are used to train this model.

[0895] Next, a trained artificial intelligence model is used to generate personalized product recommendations for the user. This process takes into account the user's past browsing and purchase history, emotional state, and market trend information. The generated product recommendations are sent to the user's smartphone in real time and displayed to the user through an appropriate interface.

[0896] When a user takes action regarding a suggested product (e.g., clicking on a product, purchasing it, or posting a review), detailed response data is collected again and sent to the server. This response data is used to understand new preferences, behavioral patterns, and emotional patterns, and the AI ​​model is further updated and improved.

[0897] As a concrete example, suppose a user is browsing camping equipment and looking for a new tent. In this case, the system collects the user's browsing history, purchase history, reviews, and sentiment data. On the server side, an artificial intelligence model is trained using this data to generate a list of optimal camping equipment for the user, who is determined to prefer "outdoor activities." This list is sent to the smartphone and displayed through the user interface. If the user purchases the suggested products and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model.

[0898] Here are some concrete examples of prompt statements:

[0899] "When users write reviews about camping equipment, use review comments and facial expression data to determine their emotional state."

[0900] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0901] Step 1:

[0902] When a user browses an e-commerce site using their smartphone, the smartphone device collects operational data. This data includes the ID of the product the user viewed, the viewing time, and the number of clicks. This data is collected and stored in a local database.

[0903] Step 2:

[0904] When users write product reviews, their smartphone cameras and text input are used to collect emotional data. The input includes user facial expressions and review comments. The OpenCV library is used to analyze the facial expressions and infer the user's emotional state. Simultaneously, the NLTK library is used to analyze the review comments and recognize the emotions expressed in the text. This emotional data is also stored in a local database.

[0905] Step 3:

[0906] The smartphone device converts collected behavioral and emotional data into a specified data format and sends it to the server via the internet. The input for this process is all behavioral and emotional data stored locally. The output is the transmission process of this data. The data transmission is performed securely using the HTTPS protocol.

[0907] Step 4:

[0908] The server stores the received behavioral and sentimental data in a database such as Amazon RDS and performs initial processing using a Python script. The input consists of behavioral and sentimental data sent from the user's terminal. Initial processing includes data cleansing and formatting. The output is the cleaned data.

[0909] Step 5:

[0910] The server uses the cleansed data to train an artificial intelligence model. The inputs include cleansed behavioral data and sentiment data. A machine learning model is built and trained using tools such as Scikit-learn's Random Forest. The output is the trained AI model.

[0911] Step 6:

[0912] Based on a trained AI model, the server generates personalized product recommendations. Inputs include the trained AI model and newly submitted user behavior and sentiment data. Market trend information is also incorporated. The AI ​​model generates an optimized product list. The output is a product recommendation list.

[0913] Step 7:

[0914] The server sends the generated product suggestion list to the user's smartphone. The input is the product suggestion list generated by the AI ​​model. The output is the product suggestion displayed on the user's smartphone.

[0915] Step 8:

[0916] Action and sentiment data are collected again when the user purchases or clicks on one of the suggested products. The input is the user's new behavior and sentiment data. This data is again stored in the local database and prepared to be sent to the server.

[0917] Step 9:

[0918] The server receives new user behavior and emotion data and updates the AI ​​model. The input is the latest behavior and emotion data from the user. The AI ​​model is retrained using this new data, enabling it to make more accurate product recommendations. The output is the updated AI model.

[0919] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0920] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio 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 specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0922] [Third Embodiment]

[0923] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0924] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0925] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0926] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0927] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the 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 image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0929] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0930] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0931] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0932] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0933] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0934] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0935] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components.

[0936] 1. Collecting user behavior data

[0937] User terminal operation

[0938] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[0939] 2. Data transmission and storage

[0940] User terminal operation

[0941] The collected behavioral data is converted into a format and sent to the server. For example, it may include information such as the frequency of access to a specific product category and the frequency of purchases.

[0942] Server operation

[0943] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[0944] 3. Training of artificial intelligence models

[0945] Server operation

[0946] Using cleansed data, an artificial intelligence model is trained to learn user preferences and behavioral patterns. For example, past purchase and browsing history is used to learn features that indicate a user's interest in outdoor activities.

[0947] 4. Generating Product Proposals

[0948] Server operation

[0949] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior and current trends. For example, as summer approaches, recommendations for trail running shoes and camping gear will be made.

[0950] User terminal operation

[0951] The user terminal receives the suggestion data sent from the server and displays it through the user interface.

[0952] 5. Collecting user responses and feedback

[0953] User actions

[0954] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[0955] User terminal operation

[0956] This response data is collected and sent to the server.

[0957] 6. Evaluating responses and updating artificial intelligence models

[0958] Server operation

[0959] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, the characteristics of products that users have left particularly positive reviews on are taken into consideration and reflected in subsequent recommendations.

[0960] Specific example

[0961] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed to them. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations. In this way, the entire system is designed to continuously improve itself to enhance the user experience.

[0962] The following describes the processing flow.

[0963] Step 1:

[0964] User actions

[0965] When a user accesses an e-commerce site and browses products, information about the viewed products is recorded in real time. This includes actions such as the user navigating to the product details page, checking reviews, and adding the product to their cart. Furthermore, if the user completes a purchase, that purchase information is also recorded.

[0966] Step 2:

[0967] User terminal operation

[0968] The user's terminal converts collected browsing history, purchase history, and review information into a specified data format. This data includes details such as product ID, viewing time, number of purchases, and review content. The converted data is then sent to the server.

[0969] Step 3:

[0970] Server operation

[0971] The server receives data sent from the user's terminal. The received data is saved to the database in real time. During the saving process, data cleansing is performed. This cleansing process includes supplementing missing data and correcting inconsistent data.

[0972] Step 4:

[0973] Server operation

[0974] The AI ​​model is trained using the cleansed data. This AI model extracts features from user behavior data and learns user preferences and patterns. For example, it learns that a user likes "outdoor activities" based on their past purchase and browsing history.

[0975] Step 5:

[0976] Server operation

[0977] Using a trained AI model, product recommendations are generated for the user. These recommendations combine the user's past behavior with current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[0978] Step 6:

[0979] User terminal operation

[0980] The system receives product suggestion data sent from the server and converts it into a display format. The converted product suggestion list is then displayed to the user through an easy-to-view interface. The user can then browse the suggested products and decide whether or not to purchase them.

[0981] Step 7:

[0982] User actions

[0983] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. Further data is generated when users provide reviews and feedback after purchase.

[0984] Step 8:

[0985] User terminal operation

[0986] The system collects data on user purchases, reviews, and feedback. This data is then converted back into a specified format and sent to the server. This data includes product IDs, rating scores, and review comments.

[0987] Step 9:

[0988] Server operation

[0989] The server analyzes the response data received from the user's terminal. This analysis helps to understand the user's latest preferences and behavioral patterns. For example, if a user gives a high rating to a particular brand of camping equipment, that information will be used as a basis for the analysis.

[0990] Step 10:

[0991] Server operation

[0992] The AI ​​model is retrained based on the new response data. This updates the AI ​​model to reflect the latest data, enabling more accurate product recommendations in the future. New features and patterns are learned, continuously improving the model's performance.

[0993] In this way, the program aims to improve the user experience by collecting user behavior data in real time and using AI to analyze and make suggestions.

[0994] (Example 1)

[0995] Next, we will describe Example 1. 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."

[0996] Traditional e-commerce systems have struggled to provide optimal product recommendations based on individual user preferences and behavior. In particular, the lack of mechanisms to collect and rapidly analyze user behavior data in real time for product recommendations led to a decline in the quality of the user experience. Furthermore, insufficient data cleansing processes to ensure the accuracy of collected data meant that improving the accuracy of artificial intelligence models remained a challenge.

[0997] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0998] In this invention, the server includes means for storing and cleaning collected behavioral data, means for training an artificial intelligence model that learns using the stored and cleansed behavioral data, and means for generating product suggestions based on the trained artificial intelligence model. This makes it possible to provide more accurate product suggestions in real time based on the behavioral data of individual users.

[0999] "User activity data" refers to information about all actions and behaviors that a user performs on a website or application. This includes browsing history, purchase history, feedback, and access frequency.

[1000] A "server" refers to a computer system that receives, stores, and processes data sent from user terminals, trains artificial intelligence models, and generates product suggestions.

[1001] "Means of collection" refers to methods and devices for capturing and recording user behavior data. This includes data capture by browser scripts and mobile applications.

[1002] "Means of transmission" refers to the means of transferring collected behavioral data from the user's terminal to the server. This includes HTTP requests and API calls.

[1003] "Means for saving and cleansing data" refers to means of saving the transmitted data to a database and performing data cleansing such as imputing missing values ​​and checking data types.

[1004] An "artificial intelligence model" refers to an algorithm or system that learns user behavior patterns and preferences through machine learning or deep learning based on data.

[1005] "Training methods" refer to the methods and devices used to train artificial intelligence models using cleansed data. This includes machine learning frameworks (e.g., TensorFlow and PyTorch).

[1006] "Methods for generating product suggestions" refers to methods that use a trained artificial intelligence model to select the most suitable products for the user and create suggestions.

[1007] "Means of displaying through a user interface" refers to methods and devices for visually presenting generated product suggestions to the user. This includes user interfaces for web pages and mobile applications.

[1008] "Response data" refers to data such as user feedback, purchase history, and ratings regarding suggested products.

[1009] "Means of analysis and updating" refers to methods of analyzing collected response data and retraining an artificial intelligence model based on the results to improve its accuracy.

[1010] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components and processing steps.

[1011] User behavior data collection

[1012] User actions

[1013] User behavioral data is collected when users browse, purchase, and provide feedback on products on e-commerce sites. If a user frequently browses a specific product category (e.g., sporting goods) and purchases related products, that data is collected in real time. Browser scripts and mobile applications are used for this purpose.

[1014] Sending and storing data

[1015] User terminal operation

[1016] The collected behavioral data is converted to JSON format and sent to the server via an HTTP request. For example, it may include logs of when a user frequently viewed a particular product category and information about products they purchased.

[1017] Server operation

[1018] The server stores the received data in a database such as MongoDB or MySQL. The stored data is then cleansed to maintain data accuracy, including imputation of missing values ​​and data type checks.

[1019] Training of artificial intelligence models

[1020] Server operation

[1021] Based on the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow and PyTorch. For example, features are extracted from a user's past purchase and browsing history to determine if the user is interested in outdoor activities.

[1022] Product Proposal Generation

[1023] Server operation

[1024] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior data and current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[1025] User terminal operation

[1026] The user terminal receives the suggestion data sent from the server and displays it through the user interface. For example, a product list might be displayed in the format, "Here are some camping items we recommend for you."

[1027] Collecting user responses and feedback

[1028] User actions

[1029] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[1030] User terminal operation

[1031] This response data is collected and sent to the server. For example, it may include data such as "I purchased this product" or review information such as "I gave it a 5-star rating."

[1032] Response evaluation and AI model updates

[1033] Server operation

[1034] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, it learns the characteristics of products that users have left particularly positive reviews on and incorporates this into subsequent recommendations.

[1035] Specific example

[1036] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed through the user interface. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations.

[1037] Examples of prompts to input into a generative AI model

[1038] "Generate suggestions recommending new camping equipment, scheduled for release next month, to users who have previously purchased outdoor gear."

[1039] "Create a list of optimal product suggestions based on the sports equipment that users frequently view."

[1040] Thus, the present invention is a system that uses user behavior data to make optimal product recommendations and improve the user experience.

[1041] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1042] Step 1:

[1043] User behavior data collection

[1044] User actions

[1045] Users browse e-commerce sites, purchase products, and provide feedback. This behavioral data is captured by each user's device (browser or mobile application). For example, if a user browses products in the "camping equipment" category for more than 10 minutes and adds a specific product to their cart, that information will be collected.

[1046] input

[1047] User browsing history, purchase history, and feedback information

[1048] output

[1049] Local cache of collected behavioral data

[1050] Step 2:

[1051] Sending data

[1052] User terminal operation

[1053] Behavioral data stored in the local cache is converted to JSON format and sent to the server using an HTTP request. For example, JSON data containing access frequency and purchase information for a specific category is generated and sent to the server.

[1054] input

[1055] Local cache behavioral data

[1056] output

[1057] Action data in JSON format sent to the server

[1058] Step 3:

[1059] Data storage and cleansing process

[1060] Server operation

[1061] The server saves the received JSON data to a database (e.g., MongoDB or MySQL). Afterward, a cleansing process is performed, including data imputation and data type checks. For example, if there are missing values ​​in the purchase history, they will be appropriately imputed.

[1062] input

[1063] Behavioral data in JSON format that was sent.

[1064] output

[1065] Cleansed database behavioral data

[1066] Step 4:

[1067] Training of artificial intelligence models

[1068] Server operation

[1069] Using the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow or PyTorch. Here, behavioral patterns and preferences are learned based on the user's past purchase and browsing history. For example, the pattern of a user frequently purchasing "outdoor equipment" is learned.

[1070] input

[1071] Cleansed behavioral data

[1072] output

[1073] Trained artificial intelligence model

[1074] Step 5:

[1075] Product Proposal Generation

[1076] Server operation

[1077] Using a trained model, the system generates optimal product recommendations based on current user data (such as recent browsing history and purchase information). For example, it might suggest new camping equipment to a user interested in "outdoor goods."

[1078] input

[1079] Current user data, trained artificial intelligence model

[1080] output

[1081] Generated product suggestion data (e.g., JSON format)

[1082] Step 6:

[1083] Product suggestion display

[1084] User terminal operation

[1085] The system analyzes product suggestion data received from the server and displays it through the user interface. For example, it can display a list of "recommended camping gear" on a webpage or mobile app screen.

[1086] input

[1087] Product proposal data sent from the server

[1088] output

[1089] Product suggestions displayed to the user

[1090] Step 7:

[1091] Collecting user responses

[1092] User actions

[1093] Users purchase suggested products, write reviews, and provide feedback. These actions are captured as response data.

[1094] input

[1095] User actions (purchase, review, feedback)

[1096] output

[1097] Local cache of collected response data

[1098] Step 8:

[1099] Sending and analyzing response data

[1100] User terminal operation

[1101] The collected response data is sent to the server. The data is then converted back to JSON format and sent via an HTTP request.

[1102] Server operation

[1103] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, improving the accuracy of future suggestions. For example, the characteristics of products that received high ratings can be reflected in future suggestions.

[1104] input

[1105] Response data sent from the user terminal

[1106] output

[1107] AI model updated based on feedback

[1108] (Application Example 1)

[1109] Next, we will explain Application Example 1. In the following explanation, 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."

[1110] Traditional e-commerce systems collect user behavior data to suggest products, but they have shortcomings in the accuracy and timeliness of these suggestions. Furthermore, it was difficult to evaluate in real time how beneficial the suggested products were to the user and incorporate that feedback into future suggestions. Additionally, the lack of smooth integration between product suggestions and the purchase process within the virtual environment resulted in a limited user experience.

[1111] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1112] In this invention, the server includes means for collecting user behavior data, means for transmitting the collected behavior data to the server, means for storing and initial processing the transmitted behavior data, means for training an artificial intelligence model that learns using the stored and initial processed behavior data, means for generating product suggestions based on the trained artificial intelligence model, means for displaying the generated product suggestions to the user, means for collecting and transmitting response data from the user to the server, means for evaluating the collected response data and updating the artificial intelligence model, and means for displaying product suggestions via a visual device used by the user in a virtual environment, enabling the selection and purchase of suggested products. This enables highly accurate and timely product suggestions based on user behavior data, and provides a seamless product selection and purchase experience in a virtual environment.

[1113] "Action data" refers to data about a user's actions and choices within an online or virtual environment, such as browsing history, purchase history, and feedback information.

[1114] A "server" refers to a computer system that receives, stores, and processes operational data over a network, and is used for training artificial intelligence models and generating product suggestions.

[1115] An "artificial intelligence model" refers to software that includes algorithms that analyze user preferences and behavior based on trained data and provide optimal product recommendations.

[1116] "Product recommendations" refer to information that indicates products or services recommended to the user, based on collected behavioral data and analysis results from artificial intelligence models.

[1117] "Visual devices" refer to devices that users wear to display information in a virtual environment, such as smart glasses and head-mounted displays.

[1118] "Response data" refers to information that shows users' reactions and feedback to suggested products, such as whether they purchased the product or wrote an article or review.

[1119] "Training" refers to the process of using collected and initially processed behavioral data to train an artificial intelligence model, thereby improving its prediction and suggestion accuracy.

[1120] A "virtual environment" refers to a virtual commercial space or store that a user experiences through a visual device using virtual reality technology.

[1121] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system improves the user experience through the following steps.

[1122] The system first collects user behavior data. This is done using visual devices such as smart glasses or head-mounted displays. These devices collect the user's gaze and gaze duration in real time, recording what the user is looking at.

[1123] The collected motion data is sent from the visual device to the server. The server stores the received data in a database and performs initial processing. This initial processing includes data cleansing, for example, removing noise from the gaze data and extracting accurate motion patterns.

[1124] Next, the server uses the cleansed data to train an artificial intelligence model. This training utilizes past purchase and browsing history. The trained AI model learns the user's preferences and behavioral patterns and generates optimal product recommendations.

[1125] Once product suggestions are generated, the server sends them to the user's visual device for display. The user can then review, select, and purchase the suggested products through their visual device. This allows the user to have a seamless product selection and purchase experience within the virtual environment.

[1126] Additionally, user response data is collected and sent to the server. This response data indicates how the user reacted to the suggested products. For example, it includes information such as whether the user purchased the product or wrote a review.

[1127] The server analyzes the received response data and provides the results as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, resulting in more accurate suggestions in the future.

[1128] As a concrete example, consider a scenario where a user wears smart glasses and walks around a virtual store, and optimal camping equipment is suggested in real time based on past behavioral data. If the user reviews the suggested products and indicates a purchase intention, that data is sent back to the server and used to improve the accuracy of the model.

[1129] An example of a prompt message is, "Based on the user's past purchase data and browsing history, please display recommended products in the virtual store on the smart glasses in real time."

[1130] In this way, the entire system can improve the user experience by providing highly accurate and timely product recommendations based on user behavior data, and by offering a seamless product selection and purchase experience within a virtual environment.

[1131] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1132] Step 1:

[1133] Collect user behavior data.

[1134] A user is wearing smart glasses and walking around a virtual store. The smart glasses collect real-time information about the user's gaze, gaze duration, and products displayed on the lenses. The input is the user's visual information, and the output is motion data containing this information. This data is temporarily stored within the visual device.

[1135] Step 2:

[1136] Send the collected motion data to the server.

[1137] The visual device collects motion data and sends it to a server. A network connection is used for this transmission. The input is motion data stored in the smart glasses, and the output is motion data that reaches the server. This data transmission is performed using the HTTPS protocol.

[1138] Step 3:

[1139] Save and initialize operation data.

[1140] The server receives operational data, saves it to a database, and performs initial processing. This initial processing includes data cleansing, such as removing noise and filling in missing data. The input is unprocessed operational data, and the output is cleansed and refined operational data.

[1141] Step 4:

[1142] Training an artificial intelligence model

[1143] The server uses pre-processed behavioral data to train an artificial intelligence model. This model learns user preferences based on past purchase and browsing history. The input is cleansed behavioral data, and the output is a trained AI model. Specifically, data analysis algorithms are used to extract important features and improve the model's prediction accuracy.

[1144] Step 5:

[1145] Generate product proposals

[1146] Product suggestions are generated based on a trained artificial intelligence model. The server considers the user's past behavior and current trend information to select the most suitable product for the user. The input consists of a trained AI model and current trend data, and the output is a list of specific product suggestions. The suggestion list includes descriptions and price information for each product.

[1147] Step 6:

[1148] Display product suggestions to users.

[1149] The generated product suggestions are sent from the server to the visual device and displayed to the user in real time. The input is product suggestion data sent from the server, and the output is a list of products displayed on the visual device's screen. The user can review the products and explore those of interest in more detail.

[1150] Step 7:

[1151] Collect response data from users and send it to the server.

[1152] When a user selects or purchases a suggested product, the visual device collects response data and sends it to a server. The input is user interaction information, and the output is response data sent to the server. This data includes purchase history and product reviews.

[1153] Step 8:

[1154] Evaluate response data and update the artificial intelligence model.

[1155] The server analyzes the received response data and feeds the results back into the artificial intelligence model. This allows the model to retrain, resulting in more accurate suggestions for the next time. The input is newly collected response data, and the output is an updated artificial intelligence model. Specifically, the product list is adjusted with an emphasis on positive feedback.

[1156] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1157] This invention is a system that collects user behavioral and emotional data and generates personalized product recommendations based on that data. The system consists of the following main components:

[1158] 1. Collecting user behavior data

[1159] User terminal operation

[1160] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[1161] 2. Collecting user sentiment data

[1162] User terminal operation

[1163] We collect user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, we analyze the emotion from that text. We also use cameras to infer emotions from the user's facial expressions and voice.

[1164] 3. Data transmission and storage

[1165] User terminal operation

[1166] The collected behavioral and emotional data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and emotional state.

[1167] Server operation

[1168] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[1169] 4. Training of the artificial intelligence model

[1170] Server operation

[1171] Using cleansed behavioral and emotional data, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, it learns characteristics that indicate a user prefers "outdoor activities" based on past purchase history, browsing history, and emotional states.

[1172] 5. Generating Product Proposals

[1173] Server operation

[1174] Based on a pre-trained artificial intelligence model, product recommendations are generated for the user. These recommendations take into account the user's behavioral history, emotional data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment might be suggested.

[1175] User terminal operation

[1176] The system receives product data suggested by the server, converts it to a display format, and displays it through the user interface.

[1177] 6. Collecting user responses and feedback

[1178] User actions

[1179] Users can browse suggested products, click on items they're interested in to see more details, add them to their cart, and complete the purchase. After the purchase, they can provide reviews and feedback, generating further data.

[1180] User terminal operation

[1181] This response data is collected, converted to a specified format, and sent to the server. This data includes product ID, rating score, review comments, sentiment status, and more.

[1182] 7. Evaluating responses and updating artificial intelligence models

[1183] Server operation

[1184] The server analyzes response data received from user terminals to understand new preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1185] Server operation

[1186] By retraining the AI ​​model based on new response data, the model's accuracy improves, and the next suggestions become even more optimized.

[1187] Specific example

[1188] For example, if a user who frequently browses and purchases camping equipment is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and sentiment data. The server trains an artificial intelligence model based on this data and generates an optimal list of camping equipment, taking into account the user's enjoyment of "outdoor activities." This list is sent to the user's terminal and displayed through the user interface. If the user purchases the suggested products and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model, further optimizing future suggestions for the user. In this way, the entire system achieves optimal product recommendations while continuously improving in real time based on the user's behavioral and sentiment data.

[1189] The following describes the processing flow.

[1190] Step 1:

[1191] User actions

[1192] A user accesses an e-commerce site and browses products. During this process, information about the viewed products is recorded in real time. For example, a user might view a sports equipment page multiple times and check the details of a specific product.

[1193] Step 2:

[1194] User terminal operation

[1195] The user's terminal converts the collected browsing history into a specified data format. This data includes product ID, browsing time, and page transition order. The converted data is then sent to the server.

[1196] Step 3:

[1197] User terminal operation

[1198] Users write reviews for specific products. Text analysis technology is used to extract sentiment data from these reviews. For example, emotions such as "satisfied" or "dissatisfied" are identified based on keywords and sentence tone used in the reviews.

[1199] Step 4:

[1200] User terminal operation

[1201] The system uses a built-in camera and microphone to collect emotional data from the user's facial expressions and voice while they are browsing products or writing reviews. For example, facial recognition technology is used to determine whether the user is smiling, angry, or otherwise in a negative mood.

[1202] Step 5:

[1203] User terminal operation

[1204] The collected behavioral and emotional data is converted into a specified data format and sent to the server. The converted data includes product ID, viewing time, number of purchases, review content, and emotional state.

[1205] Step 6:

[1206] Server operation

[1207] The server receives data sent from the user's terminal. The received data is saved to the database in real time. During the saving process, data cleansing is performed, correcting inconsistent data and supplementing missing data.

[1208] Step 7:

[1209] Server operation

[1210] Using cleansed behavioral and emotional data, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, it learns characteristics that identify a user as being interested in "outdoor activities" based on past purchase history, browsing history, and emotional state.

[1211] Step 8:

[1212] Server operation

[1213] Based on a pre-trained artificial intelligence model, product recommendations are generated for the user. These recommendations take into account the user's behavioral history, emotional data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment might be suggested.

[1214] Step 9:

[1215] User terminal operation

[1216] The system receives product suggestion data sent from the server, converts it into a display format, and displays it through the user interface. Users can review the product suggestions and view details through this interface.

[1217] Step 10:

[1218] User actions

[1219] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. After purchase, they are also required to provide reviews and feedback, documenting their feelings about the product.

[1220] Step 11:

[1221] User terminal operation

[1222] The system collects data on user purchases, reviews, and feedback. This data is converted to a specified format and sent to the server. The data includes product ID, rating score, review comments, and sentiment status.

[1223] Step 12:

[1224] Server operation

[1225] The server analyzes the response data received from the user's terminal. This analysis reveals the user's latest preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1226] Step 13:

[1227] Server operation

[1228] The AI ​​model is retrained based on the new response data. This retraining improves the accuracy of the AI ​​model, further optimizing future suggestions. New features and patterns are learned, continuously improving the model's performance.

[1229] In this way, the program collects user behavior and emotional data in real time, and uses AI to analyze and make suggestions, thereby improving the user experience.

[1230] (Example 2)

[1231] Next, we will describe Example 2. 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."

[1232] Traditional e-commerce systems only collect user behavior data to suggest products, making it difficult to provide personalized recommendations based on user emotions and interests. Furthermore, the accuracy of product recommendations was low, making it difficult to improve user satisfaction. In addition, because user emotional data was not reflected in the system, more appropriate product recommendations could not be achieved.

[1233] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1234] In this invention, the server includes means for collecting user behavior data and emotional data, means for transmitting the collected behavior data and emotional data to the server, and means for storing and initial processing the transmitted behavior data and emotional data. This enables more personalized product recommendations by utilizing both the user's behavior data and emotional data. Specifically, it includes means for collecting emotional data in real time using user emotion recognition technology, means for inferring emotions from the user's facial expressions and voice using a camera, and means for improving the accuracy of product recommendations using the user's emotional data, thereby realizing more accurate product recommendations.

[1235] "Action data" refers to information about user actions such as browsing, purchasing, and providing feedback on e-commerce sites.

[1236] "Emotional data" refers to information about a user's emotional state obtained from facial recognition technology, text analysis, and facial expressions and voice captured using a camera.

[1237] "Means of collection" refers to hardware and software for acquiring user behavioral data and emotional data in real time.

[1238] "Means of transmission" refers to communication means for converting collected behavioral data and emotional data into the required format and sending it to the server.

[1239] "Means for saving and initial processing" refers to the process of saving the behavioral and emotional data sent to the server into a database, verifying the integrity of the data, and cleansing up any inaccurate data.

[1240] "Training methods" refer to methods for building and training artificial intelligence models that learn user preferences, behavioral patterns, and emotional patterns using stored and pre-processed data.

[1241] "Methods for generating product suggestions" refers to methods that utilize trained artificial intelligence models to generate personalized and optimal product suggestions for each user.

[1242] "Means of display" refers to a mechanism for sending generated product suggestions to the user's terminal and presenting them visually through the user interface.

[1243] "Response data" refers to information about actions users take in response to product suggestions (e.g., viewing products, purchasing them, writing reviews).

[1244] "Methods for updating" refer to methods of analyzing collected response data and retraining the artificial intelligence model to improve its accuracy.

[1245] "Emotion recognition technology" is a technology that uses facial recognition and text analysis of users, as well as a camera to analyze facial expressions and voice, in order to infer the user's emotional state.

[1246] This invention is a system that collects user behavior data and emotional data and generates product recommendations that are optimal for each individual user based on that data. This system includes the following main components:

[1247] 1. Collecting user behavior data

[1248] This refers to the user's terminal activity. This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[1249] 2. Collecting user sentiment data

[1250] This involves the operation of the user's terminal. It collects user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, the emotions are analyzed from the text. It also uses the camera to infer emotions from the user's facial expressions and voice. Specifically, emotion recognition software (e.g., Microsoft Azure Face API) is used.

[1251] 3. Data transmission and storage

[1252] This refers to the user's terminal activity. The collected activity and sentiment data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and sentiment state.

[1253] This describes the server's operation. The server stores the received data in a database (e.g., Amazon RDS) in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[1254] 4. Training of the artificial intelligence model

[1255] This is the server operation. Using cleansed behavioral and emotional data, it trains an artificial intelligence model that learns user preferences, behavioral patterns, and emotional patterns. For example, it runs a training script written in Python and builds the model using TensorFlow. From past purchase and browsing history, it learns that the user prefers "outdoor activities."

[1256] 5. Generating Product Proposals

[1257] This describes the server's operation. Based on a pre-trained artificial intelligence model, it generates product suggestions for the user. Here, the suggestions are made considering the user's behavioral history, sentiment data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment may be suggested.

[1258] This describes the operation of the user terminal. It receives product data proposed from the server, converts it into a display format, and displays it through the user interface.

[1259] 6. Collecting user responses and feedback

[1260] This describes user behavior. Users browse suggested products, click on items of interest to view details, add them to their cart, and complete the purchase. Further data is generated when users provide reviews and feedback after the purchase.

[1261] This describes the user terminal's actions. It collects this response data, converts it to a specified format, and sends it to the server. This data includes product ID, rating score, review comments, and sentiment status.

[1262] 7. Evaluating responses and updating artificial intelligence models

[1263] This is how the server works. The server analyzes response data received from the user's terminal to understand new preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1264] This is a server operation. By retraining the AI ​​model based on new response data, the model's accuracy improves, and the next suggestions become even more optimized.

[1265] Specific example

[1266] For example, if a user who frequently browses and purchases camping gear is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and sentiment data. The server uses this data to train an artificial intelligence model and generates an optimal list of camping gear, taking into account the user's enjoyment of "outdoor activities." This list is sent to the user's terminal and displayed through the user interface. If the user purchases the suggested items and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model, further optimizing future suggestions for the user.

[1267] Example of a prompt

[1268] "Please suggest the latest tents to users who frequently browse and purchase camping equipment."

[1269] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1270] Step 1:

[1271] User behavior data collection

[1272] User terminal operation

[1273] We collect this behavioral data when users browse, purchase, and provide feedback on products on e-commerce sites. For example, when a user is browsing a specific sporting item, we record the product ID, the start time of the browsing, and the end time. If the user completes a purchase, we also collect the purchased product ID, the date and time of purchase, and the purchase amount.

[1274] Input: User's behavior on the website

[1275] Output: Operational data such as product ID, viewing time, and purchase data.

[1276] Step 2:

[1277] Collecting user sentiment data

[1278] User terminal operation

[1279] Using facial recognition technology and text analysis, we collect emotional data from users when they write product reviews or use the website. Specifically, we extract emotions from the text of product reviews written by users, and infer emotions from facial expressions and voice using a camera. For this purpose, we use emotion recognition software.

[1280] Input: User review text, camera footage, audio

[1281] Output: Emotional data such as type of emotion (positive, negative, etc.) and intensity.

[1282] Step 3:

[1283] Sending and storing data

[1284] User terminal operation

[1285] The collected behavioral and emotional data is converted into JSON format and sent to the server in real time. For example, the dataset might look like this: "Product ID: 1234, Viewing time: 5 minutes, Purchase date and time: 2023-10-12, Review emotion: Positive, Facial emotion: Smile".

[1286] Input: Behavioral data, emotional data

[1287] Output: Data in JSON format

[1288] Server operation

[1289] The server stores the received data in a database and performs a cleansing process as an initial step. This removes inconsistent or missing data to ensure data integrity.

[1290] Input: Data in JSON format

[1291] Output: Cleansed and consistent data

[1292] Step 4:

[1293] Training of artificial intelligence models

[1294] Server operation

[1295] Based on cleansed behavioral and emotional data, an AI model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, a script written in Python is executed, and the model is built using TensorFlow. The training data includes past purchase history, browsing history, and emotional states.

[1296] Input: Data after cleansing

[1297] Output: Trained AI model

[1298] Step 5:

[1299] Product Proposal Generation

[1300] Server operation

[1301] Using a pre-trained artificial intelligence model, the system generates personalized product recommendations for each user. Specifically, it considers the user's past behavioral history, emotional data, and current trend information when making recommendations. For example, "as summer approaches, it might suggest trail running shoes or camping equipment."

[1302] Input: Trained AI model, user behavior data, sentiment data, trend information

[1303] Output: Product Proposal List

[1304] User terminal operation

[1305] The system receives product data suggested by the server and converts it into a format for display in the user interface. This allows the user to visually see product images and detailed information.

[1306] Input: Product Proposal List

[1307] Output: Product list displayed in the user interface

[1308] Step 6:

[1309] Collecting user responses and feedback

[1310] User actions

[1311] Users can view suggested products, check their details, add them to their cart, and complete the purchase. After purchase, they can provide reviews and feedback.

[1312] Input: User actions (browsing, purchasing, reviewing)

[1313] Output: Response data

[1314] User terminal operation

[1315] The collected response data is converted to JSON format and sent to the server. This data includes product ID, rating score, review comments, sentiment status, etc.

[1316] Input: User action data

[1317] Output: Response data in JSON format

[1318] Step 7:

[1319] Response evaluation and AI model updates

[1320] Server operation

[1321] The server analyzes the received response data to understand new user preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1322] Input: Response data in JSON format

[1323] Output: Updated user preference data, behavioral pattern data, emotional pattern data

[1324] Server operation

[1325] The AI ​​model is retrained based on the new response data to improve its accuracy. This will further optimize future suggestions.

[1326] Input: Updated user preference data, behavioral pattern data, emotional pattern data

[1327] Output: Improved AI model

[1328] (Application Example 2)

[1329] Next, we will explain application example 2. In the following explanation, 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."

[1330] Traditional e-commerce systems suggested products based on users' purchase and browsing history, but it was difficult to provide optimal product suggestions that took into account user emotions and real-time behavior. Furthermore, the lack of functionality to collect and analyze user emotion data prevented personalized suggestions that reflected user feelings. As a result, improvements in the user experience were limited.

[1331] The specific processing performed 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 user behavior data and emotion data, means for transmitting the collected behavior data and emotion data to the server, and means for storing and initial processing the transmitted behavior data and emotion data. This makes it possible to provide more precise personalized product suggestions based on the user's behavior and emotions.

[1332] "Action data" refers to the user's behavioral history on the e-commerce system, including purchase history, browsing history, and feedback information.

[1333] "Emotional data" refers to information about a user's emotional state obtained through methods such as facial expressions, voice, and text analysis.

[1334] "Personalized product recommendations" refer to product recommendations optimized for each user, generated based on their individual behavioral and emotional data.

[1335] "Facial recognition" refers to a technology that captures a user's facial expressions and analyzes their emotions from those expressions.

[1336] "Text analysis" refers to the technology that analyzes text data entered by users and understands their emotions and intentions from its content.

[1337] "Trend information" refers to data about current market and consumer trends.

[1338] "Response data" refers to feedback information obtained from user actions taken in response to product suggestions (e.g., clicks, purchases, reviews, etc.).

[1339] An "artificial intelligence model" refers to a machine learning model that learns from collected and analyzed data to recognize and predict patterns and trends.

[1340] "Real-time" refers to processing and making suggestions that immediately reflect changes in user behavior and emotions.

[1341] This invention relates to a system that collects user behavior data and emotional data and generates personalized product recommendations based on that data. This system is primarily implemented using the following hardware and software components.

[1342] First, the user's smartphone will have programs installed to collect behavioral and emotional data. Collecting behavioral data includes functions to acquire information (purchase history, browsing history, feedback information, etc.) when the user browses or purchases products on e-commerce sites. Simultaneously, emotional data will be collected using facial recognition and text analysis software utilizing the smartphone's camera. Specifically, the OpenCV library will be used to capture the user's facial expressions and analyze their emotions in real time. Text analysis will use the NLTK library to analyze emotions from text data such as reviews and feedback.

[1343] The collected behavioral and emotional data are converted to a specified data format and sent to an AWS (Amazon Web Services) server via the internet. The server stores the received data in a database such as Amazon RDS and performs data cleansing using a Python script. After the data is cleansed, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. Machine learning algorithms such as Scikit-learn's Random Forest Classifier are used to train this model.

[1344] Next, a trained artificial intelligence model is used to generate personalized product recommendations for the user. This process takes into account the user's past browsing and purchase history, emotional state, and market trend information. The generated product recommendations are sent to the user's smartphone in real time and displayed to the user through an appropriate interface.

[1345] When a user takes action regarding a suggested product (e.g., clicking on a product, purchasing it, or posting a review), detailed response data is collected again and sent to the server. This response data is used to understand new preferences, behavioral patterns, and emotional patterns, and the AI ​​model is further updated and improved.

[1346] As a concrete example, suppose a user is browsing camping equipment and looking for a new tent. In this case, the system collects the user's browsing history, purchase history, reviews, and sentiment data. On the server side, an artificial intelligence model is trained using this data to generate a list of optimal camping equipment for the user, who is determined to prefer "outdoor activities." This list is sent to the smartphone and displayed through the user interface. If the user purchases the suggested products and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model.

[1347] Here are some concrete examples of prompt statements:

[1348] "When users write reviews about camping equipment, use review comments and facial expression data to determine their emotional state."

[1349] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1350] Step 1:

[1351] When a user browses an e-commerce site using their smartphone, the smartphone device collects operational data. This data includes the ID of the product the user viewed, the viewing time, and the number of clicks. This data is collected and stored in a local database.

[1352] Step 2:

[1353] When users write product reviews, their smartphone cameras and text input are used to collect emotional data. The input includes user facial expressions and review comments. The OpenCV library is used to analyze the facial expressions and infer the user's emotional state. Simultaneously, the NLTK library is used to analyze the review comments and recognize the emotions expressed in the text. This emotional data is also stored in a local database.

[1354] Step 3:

[1355] The smartphone device converts collected behavioral and emotional data into a specified data format and sends it to the server via the internet. The input for this process is all behavioral and emotional data stored locally. The output is the transmission process of this data. The data transmission is performed securely using the HTTPS protocol.

[1356] Step 4:

[1357] The server stores the received behavioral and sentimental data in a database such as Amazon RDS and performs initial processing using a Python script. The input consists of behavioral and sentimental data sent from the user's terminal. Initial processing includes data cleansing and formatting. The output is the cleaned data.

[1358] Step 5:

[1359] The server uses the cleansed data to train an artificial intelligence model. The inputs include cleansed behavioral data and sentiment data. A machine learning model is built and trained using tools such as Scikit-learn's Random Forest. The output is the trained AI model.

[1360] Step 6:

[1361] Based on a trained AI model, the server generates personalized product recommendations. Inputs include the trained AI model and newly submitted user behavior and sentiment data. Market trend information is also incorporated. The AI ​​model generates an optimized product list. The output is a product recommendation list.

[1362] Step 7:

[1363] The server sends the generated product suggestion list to the user's smartphone. The input is the product suggestion list generated by the AI ​​model. The output is the product suggestion displayed on the user's smartphone.

[1364] Step 8:

[1365] Action and sentiment data are collected again when the user purchases or clicks on one of the suggested products. The input is the user's new behavior and sentiment data. This data is again stored in the local database and prepared to be sent to the server.

[1366] Step 9:

[1367] The server receives new user behavior and emotion data and updates the AI ​​model. The input is the latest behavior and emotion data from the user. The AI ​​model is retrained using this new data, enabling it to make more accurate product recommendations. The output is the updated AI model.

[1368] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1369] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1370] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1371] [Fourth Embodiment]

[1372] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1373] As shown in Figure 7, the 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.

[1374] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1375] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1376] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1377] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1378] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1379] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive 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 robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1380] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1381] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[1382] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1383] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1384] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1385] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components.

[1386] 1. Collecting user behavior data

[1387] User terminal operation

[1388] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[1389] 2. Data transmission and storage

[1390] User terminal operation

[1391] The collected behavioral data is converted into a format and sent to the server. For example, it may include information such as the frequency of access to a specific product category and the frequency of purchases.

[1392] Server operation

[1393] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[1394] 3. Training of artificial intelligence models

[1395] Server operation

[1396] Using cleansed data, an artificial intelligence model is trained to learn user preferences and behavioral patterns. For example, past purchase and browsing history is used to learn features that indicate a user's interest in outdoor activities.

[1397] 4. Generating Product Proposals

[1398] Server operation

[1399] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior and current trends. For example, as summer approaches, recommendations for trail running shoes and camping gear will be made.

[1400] User terminal operation

[1401] The user terminal receives the suggestion data sent from the server and displays it through the user interface.

[1402] 5. Collecting user responses and feedback

[1403] User actions

[1404] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[1405] User terminal operation

[1406] This response data is collected and sent to the server.

[1407] 6. Evaluating responses and updating artificial intelligence models

[1408] Server operation

[1409] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, the characteristics of products that users have left particularly positive reviews on are taken into consideration and reflected in subsequent recommendations.

[1410] Specific example

[1411] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed to them. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations. In this way, the entire system is designed to continuously improve itself to enhance the user experience.

[1412] The following describes the processing flow.

[1413] Step 1:

[1414] User actions

[1415] When a user accesses an e-commerce site and browses products, information about the viewed products is recorded in real time. This includes actions such as the user navigating to the product details page, checking reviews, and adding the product to their cart. Furthermore, if the user completes a purchase, that purchase information is also recorded.

[1416] Step 2:

[1417] User terminal operation

[1418] The user's terminal converts collected browsing history, purchase history, and review information into a specified data format. This data includes details such as product ID, viewing time, number of purchases, and review content. The converted data is then sent to the server.

[1419] Step 3:

[1420] Server operation

[1421] The server receives data sent from the user's terminal. The received data is saved to the database in real time. During the saving process, data cleansing is performed. This cleansing process includes supplementing missing data and correcting inconsistent data.

[1422] Step 4:

[1423] Server operation

[1424] The AI ​​model is trained using the cleansed data. This AI model extracts features from user behavior data and learns user preferences and patterns. For example, it learns that a user likes "outdoor activities" based on their past purchase and browsing history.

[1425] Step 5:

[1426] Server operation

[1427] Using a trained AI model, product recommendations are generated for the user. These recommendations combine the user's past behavior with current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[1428] Step 6:

[1429] User terminal operation

[1430] The system receives product suggestion data sent from the server and converts it into a display format. The converted product suggestion list is then displayed to the user through an easy-to-view interface. The user can then browse the suggested products and decide whether or not to purchase them.

[1431] Step 7:

[1432] User actions

[1433] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. Further data is generated when users provide reviews and feedback after purchase.

[1434] Step 8:

[1435] User terminal operation

[1436] The system collects data on user purchases, reviews, and feedback. This data is then converted back into a specified format and sent to the server. This data includes product IDs, rating scores, and review comments.

[1437] Step 9:

[1438] Server operation

[1439] The server analyzes the response data received from the user's terminal. This analysis helps to understand the user's latest preferences and behavioral patterns. For example, if a user gives a high rating to a particular brand of camping equipment, that information will be used as a basis for the analysis.

[1440] Step 10:

[1441] Server operation

[1442] The AI ​​model is retrained based on the new response data. This updates the AI ​​model to reflect the latest data, enabling more accurate product recommendations in the future. New features and patterns are learned, continuously improving the model's performance.

[1443] In this way, the program aims to improve the user experience by collecting user behavior data in real time and using AI to analyze and make suggestions.

[1444] (Example 1)

[1445] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1446] Traditional e-commerce systems have struggled to provide optimal product recommendations based on individual user preferences and behavior. In particular, the lack of mechanisms to collect and rapidly analyze user behavior data in real time for product recommendations led to a decline in the quality of the user experience. Furthermore, insufficient data cleansing processes to ensure the accuracy of collected data meant that improving the accuracy of artificial intelligence models remained a challenge.

[1447] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1448] In this invention, the server includes means for storing and cleaning collected behavioral data, means for training an artificial intelligence model that learns using the stored and cleansed behavioral data, and means for generating product suggestions based on the trained artificial intelligence model. This makes it possible to provide more accurate product suggestions in real time based on the behavioral data of individual users.

[1449] "User activity data" refers to information about all actions and behaviors that a user performs on a website or application. This includes browsing history, purchase history, feedback, and access frequency.

[1450] A "server" refers to a computer system that receives, stores, and processes data sent from user terminals, trains artificial intelligence models, and generates product suggestions.

[1451] "Means of collection" refers to methods and devices for capturing and recording user behavior data. This includes data capture by browser scripts and mobile applications.

[1452] "Means of transmission" refers to the means of transferring collected behavioral data from the user's terminal to the server. This includes HTTP requests and API calls.

[1453] "Means for saving and cleansing data" refers to means of saving the transmitted data to a database and performing data cleansing such as imputing missing values ​​and checking data types.

[1454] An "artificial intelligence model" refers to an algorithm or system that learns user behavior patterns and preferences through machine learning or deep learning based on data.

[1455] "Training methods" refer to the methods and devices used to train artificial intelligence models using cleansed data. This includes machine learning frameworks (e.g., TensorFlow and PyTorch).

[1456] "Methods for generating product suggestions" refers to methods that use a trained artificial intelligence model to select the most suitable products for the user and create suggestions.

[1457] "Means of displaying through a user interface" refers to methods and devices for visually presenting generated product suggestions to the user. This includes user interfaces for web pages and mobile applications.

[1458] "Response data" refers to data such as user feedback, purchase history, and ratings regarding suggested products.

[1459] "Means of analysis and updating" refers to methods of analyzing collected response data and retraining an artificial intelligence model based on the results to improve its accuracy.

[1460] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system consists of the following main components and processing steps.

[1461] User behavior data collection

[1462] User actions

[1463] User behavioral data is collected when users browse, purchase, and provide feedback on products on e-commerce sites. If a user frequently browses a specific product category (e.g., sporting goods) and purchases related products, that data is collected in real time. Browser scripts and mobile applications are used for this purpose.

[1464] Sending and storing data

[1465] User terminal operation

[1466] The collected behavioral data is converted to JSON format and sent to the server via an HTTP request. For example, it may include logs of when a user frequently viewed a particular product category and information about products they purchased.

[1467] Server operation

[1468] The server stores the received data in a database such as MongoDB or MySQL. The stored data is then cleansed to maintain data accuracy, including imputation of missing values ​​and data type checks.

[1469] Training of artificial intelligence models

[1470] Server operation

[1471] Based on the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow and PyTorch. For example, features are extracted from a user's past purchase and browsing history to determine if the user is interested in outdoor activities.

[1472] Product Proposal Generation

[1473] Server operation

[1474] Product recommendations are generated for the user based on a pre-trained artificial intelligence model. These recommendations take into account the user's past behavior data and current trend information. For example, as summer approaches, trail running shoes and camping gear might be suggested.

[1475] User terminal operation

[1476] The user terminal receives the suggestion data sent from the server and displays it through the user interface. For example, a product list might be displayed in the format, "Here are some camping items we recommend for you."

[1477] Collecting user responses and feedback

[1478] User actions

[1479] Response data is generated when a user purchases a suggested product, writes a review, or provides feedback. For example, a user might purchase suggested camping equipment and write a review.

[1480] User terminal operation

[1481] This response data is collected and sent to the server. For example, it may include data such as "I purchased this product" or review information such as "I gave it a 5-star rating."

[1482] Response evaluation and AI model updates

[1483] Server operation

[1484] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. Through this process, the AI ​​model retrains itself, enabling more accurate product recommendations. For example, it learns the characteristics of products that users have left particularly positive reviews on and incorporates this into subsequent recommendations.

[1485] Specific example

[1486] For example, if a user interested in the outdoors is looking for new camping gear, their device sends their past search and purchase history to the server. The server uses this data to train an artificial intelligence model and generate a list of camping gear best suited to the user. This list is sent to the user's device and displayed through the user interface. When the user purchases an item from the list and posts a review, this response data is sent back to the server, updating the AI ​​model and further optimizing future recommendations.

[1487] Examples of prompts to input into a generative AI model

[1488] "Generate suggestions recommending new camping equipment, scheduled for release next month, to users who have previously purchased outdoor gear."

[1489] "Create a list of optimal product suggestions based on the sports equipment that users frequently view."

[1490] Thus, the present invention is a system that uses user behavior data to make optimal product recommendations and improve the user experience.

[1491] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1492] Step 1:

[1493] User behavior data collection

[1494] User actions

[1495] Users browse e-commerce sites, purchase products, and provide feedback. This behavioral data is captured by each user's device (browser or mobile application). For example, if a user browses products in the "camping equipment" category for more than 10 minutes and adds a specific product to their cart, that information will be collected.

[1496] input

[1497] User browsing history, purchase history, and feedback information

[1498] output

[1499] Local cache of collected behavioral data

[1500] Step 2:

[1501] Sending data

[1502] User terminal operation

[1503] Behavioral data stored in the local cache is converted to JSON format and sent to the server using an HTTP request. For example, JSON data containing access frequency and purchase information for a specific category is generated and sent to the server.

[1504] input

[1505] Local cache behavioral data

[1506] output

[1507] Action data in JSON format sent to the server

[1508] Step 3:

[1509] Data storage and cleansing process

[1510] Server operation

[1511] The server saves the received JSON data to a database (e.g., MongoDB or MySQL). Afterward, a cleansing process is performed, including data imputation and data type checks. For example, if there are missing values ​​in the purchase history, they will be appropriately imputed.

[1512] input

[1513] Behavioral data in JSON format that was sent.

[1514] output

[1515] Cleansed database behavioral data

[1516] Step 4:

[1517] Training of artificial intelligence models

[1518] Server operation

[1519] Using the cleansed data, an artificial intelligence model is trained using machine learning frameworks such as TensorFlow or PyTorch. Here, behavioral patterns and preferences are learned based on the user's past purchase and browsing history. For example, the pattern of a user frequently purchasing "outdoor equipment" is learned.

[1520] input

[1521] Cleansed behavioral data

[1522] output

[1523] Trained artificial intelligence model

[1524] Step 5:

[1525] Product Proposal Generation

[1526] Server operation

[1527] Using a trained model, the system generates optimal product recommendations based on current user data (such as recent browsing history and purchase information). For example, it might suggest new camping equipment to a user interested in "outdoor goods."

[1528] input

[1529] Current user data, trained artificial intelligence model

[1530] output

[1531] Generated product suggestion data (e.g., JSON format)

[1532] Step 6:

[1533] Product suggestion display

[1534] User terminal operation

[1535] The system analyzes product suggestion data received from the server and displays it through the user interface. For example, it can display a list of "recommended camping gear" on a webpage or mobile app screen.

[1536] input

[1537] Product proposal data sent from the server

[1538] output

[1539] Product suggestions displayed to the user

[1540] Step 7:

[1541] Collecting user responses

[1542] User actions

[1543] Users purchase suggested products, write reviews, and provide feedback. These actions are captured as response data.

[1544] input

[1545] User actions (purchase, review, feedback)

[1546] output

[1547] Local cache of collected response data

[1548] Step 8:

[1549] Sending and analyzing response data

[1550] User terminal operation

[1551] The collected response data is sent to the server. The data is then converted back to JSON format and sent via an HTTP request.

[1552] Server operation

[1553] The received response data is analyzed, and the results are provided as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, improving the accuracy of future suggestions. For example, the characteristics of products that received high ratings can be reflected in future suggestions.

[1554] input

[1555] Response data sent from the user terminal

[1556] output

[1557] AI model updated based on feedback

[1558] (Application Example 1)

[1559] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1560] Traditional e-commerce systems collect user behavior data to suggest products, but they have shortcomings in the accuracy and timeliness of these suggestions. Furthermore, it was difficult to evaluate in real time how beneficial the suggested products were to the user and incorporate that feedback into future suggestions. Additionally, the lack of smooth integration between product suggestions and the purchase process within the virtual environment resulted in a limited user experience.

[1561] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1562] In this invention, the server includes means for collecting user behavior data, means for transmitting the collected behavior data to the server, means for storing and initial processing the transmitted behavior data, means for training an artificial intelligence model that learns using the stored and initial processed behavior data, means for generating product suggestions based on the trained artificial intelligence model, means for displaying the generated product suggestions to the user, means for collecting and transmitting response data from the user to the server, means for evaluating the collected response data and updating the artificial intelligence model, and means for displaying product suggestions via a visual device used by the user in a virtual environment, enabling the selection and purchase of suggested products. This enables highly accurate and timely product suggestions based on user behavior data, and provides a seamless product selection and purchase experience in a virtual environment.

[1563] "Action data" refers to data about a user's actions and choices within an online or virtual environment, such as browsing history, purchase history, and feedback information.

[1564] A "server" refers to a computer system that receives, stores, and processes operational data over a network, and is used for training artificial intelligence models and generating product suggestions.

[1565] An "artificial intelligence model" refers to software that includes algorithms that analyze user preferences and behavior based on trained data and provide optimal product recommendations.

[1566] "Product recommendations" refer to information that indicates products or services recommended to the user, based on collected behavioral data and analysis results from artificial intelligence models.

[1567] "Visual devices" refer to devices that users wear to display information in a virtual environment, such as smart glasses and head-mounted displays.

[1568] "Response data" refers to information that shows users' reactions and feedback to suggested products, such as whether they purchased the product or wrote an article or review.

[1569] "Training" refers to the process of using collected and initially processed behavioral data to train an artificial intelligence model, thereby improving its prediction and suggestion accuracy.

[1570] A "virtual environment" refers to a virtual commercial space or store that a user experiences through a visual device using virtual reality technology.

[1571] This invention is a system that collects user behavior data and generates product recommendations that are optimal for each individual user based on that data. This system improves the user experience through the following steps.

[1572] The system first collects user behavior data. This is done using visual devices such as smart glasses or head-mounted displays. These devices collect the user's gaze and gaze duration in real time, recording what the user is looking at.

[1573] The collected motion data is sent from the visual device to the server. The server stores the received data in a database and performs initial processing. This initial processing includes data cleansing, for example, removing noise from the gaze data and extracting accurate motion patterns.

[1574] Next, the server uses the cleansed data to train an artificial intelligence model. This training utilizes past purchase and browsing history. The trained AI model learns the user's preferences and behavioral patterns and generates optimal product recommendations.

[1575] Once product suggestions are generated, the server sends them to the user's visual device for display. The user can then review, select, and purchase the suggested products through their visual device. This allows the user to have a seamless product selection and purchase experience within the virtual environment.

[1576] Additionally, user response data is collected and sent to the server. This response data indicates how the user reacted to the suggested products. For example, it includes information such as whether the user purchased the product or wrote a review.

[1577] The server analyzes the received response data and provides the results as feedback to the artificial intelligence model. This allows the AI ​​model to retrain, resulting in more accurate suggestions in the future.

[1578] As a concrete example, consider a scenario where a user wears smart glasses and walks around a virtual store, and optimal camping equipment is suggested in real time based on past behavioral data. If the user reviews the suggested products and indicates a purchase intention, that data is sent back to the server and used to improve the accuracy of the model.

[1579] An example of a prompt message is, "Based on the user's past purchase data and browsing history, please display recommended products in the virtual store on the smart glasses in real time."

[1580] In this way, the entire system can improve the user experience by providing highly accurate and timely product recommendations based on user behavior data, and by offering a seamless product selection and purchase experience within a virtual environment.

[1581] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1582] Step 1:

[1583] Collect user behavior data.

[1584] A user is wearing smart glasses and walking around a virtual store. The smart glasses collect real-time information about the user's gaze, gaze duration, and products displayed on the lenses. The input is the user's visual information, and the output is motion data containing this information. This data is temporarily stored within the visual device.

[1585] Step 2:

[1586] Send the collected motion data to the server.

[1587] The visual device collects motion data and sends it to a server. A network connection is used for this transmission. The input is motion data stored in the smart glasses, and the output is motion data that reaches the server. This data transmission is performed using the HTTPS protocol.

[1588] Step 3:

[1589] Save and initialize operation data.

[1590] The server receives operational data, saves it to a database, and performs initial processing. This initial processing includes data cleansing, such as removing noise and filling in missing data. The input is unprocessed operational data, and the output is cleansed and refined operational data.

[1591] Step 4:

[1592] Training an artificial intelligence model

[1593] The server uses pre-processed behavioral data to train an artificial intelligence model. This model learns user preferences based on past purchase and browsing history. The input is cleansed behavioral data, and the output is a trained AI model. Specifically, data analysis algorithms are used to extract important features and improve the model's prediction accuracy.

[1594] Step 5:

[1595] Generate product proposals

[1596] Product suggestions are generated based on a trained artificial intelligence model. The server considers the user's past behavior and current trend information to select the most suitable product for the user. The input consists of a trained AI model and current trend data, and the output is a list of specific product suggestions. The suggestion list includes descriptions and price information for each product.

[1597] Step 6:

[1598] Display product suggestions to users.

[1599] The generated product suggestions are sent from the server to the visual device and displayed to the user in real time. The input is product suggestion data sent from the server, and the output is a list of products displayed on the visual device's screen. The user can review the products and explore those of interest in more detail.

[1600] Step 7:

[1601] Collect response data from users and send it to the server.

[1602] When a user selects or purchases a suggested product, the visual device collects response data and sends it to a server. The input is user interaction information, and the output is response data sent to the server. This data includes purchase history and product reviews.

[1603] Step 8:

[1604] Evaluate response data and update the artificial intelligence model.

[1605] The server analyzes the received response data and feeds the results back into the artificial intelligence model. This allows the model to retrain, resulting in more accurate suggestions for the next time. The input is newly collected response data, and the output is an updated artificial intelligence model. Specifically, the product list is adjusted with an emphasis on positive feedback.

[1606] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1607] This invention is a system that collects user behavioral and emotional data and generates personalized product recommendations based on that data. The system consists of the following main components:

[1608] 1. Collecting user behavior data

[1609] User terminal operation

[1610] This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[1611] 2. Collecting user sentiment data

[1612] User terminal operation

[1613] We collect user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, we analyze the emotion from that text. We also use cameras to infer emotions from the user's facial expressions and voice.

[1614] 3. Data transmission and storage

[1615] User terminal operation

[1616] The collected behavioral and emotional data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and emotional state.

[1617] Server operation

[1618] The server saves the received data to the database in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[1619] 4. Training of the artificial intelligence model

[1620] Server operation

[1621] Using cleansed behavioral and emotional data, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, it learns characteristics that indicate a user prefers "outdoor activities" based on past purchase history, browsing history, and emotional states.

[1622] 5. Generating Product Proposals

[1623] Server operation

[1624] Based on a pre-trained artificial intelligence model, product recommendations are generated for the user. These recommendations take into account the user's behavioral history, emotional data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment might be suggested.

[1625] User terminal operation

[1626] The system receives product data suggested by the server, converts it to a display format, and displays it through the user interface.

[1627] 6. Collecting user responses and feedback

[1628] User actions

[1629] Users can browse suggested products, click on items they're interested in to see more details, add them to their cart, and complete the purchase. After the purchase, they can provide reviews and feedback, generating further data.

[1630] User terminal operation

[1631] This response data is collected, converted to a specified format, and sent to the server. This data includes product ID, rating score, review comments, sentiment status, and more.

[1632] 7. Evaluating responses and updating artificial intelligence models

[1633] Server operation

[1634] The server analyzes response data received from user terminals to understand new preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1635] Server operation

[1636] By retraining the AI ​​model based on new response data, the model's accuracy improves, and the next suggestions become even more optimized.

[1637] Specific example

[1638] For example, if a user who frequently browses and purchases camping equipment is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and sentiment data. The server trains an artificial intelligence model based on this data and generates an optimal list of camping equipment, taking into account the user's enjoyment of "outdoor activities." This list is sent to the user's terminal and displayed through the user interface. If the user purchases the suggested products and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model, further optimizing future suggestions for the user. In this way, the entire system achieves optimal product recommendations while continuously improving in real time based on the user's behavioral and sentiment data.

[1639] The following describes the processing flow.

[1640] Step 1:

[1641] User actions

[1642] A user accesses an e-commerce site and browses products. During this process, information about the viewed products is recorded in real time. For example, a user might view a sports equipment page multiple times and check the details of a specific product.

[1643] Step 2:

[1644] User terminal operation

[1645] The user's terminal converts the collected browsing history into a specified data format. This data includes product ID, browsing time, and page transition order. The converted data is then sent to the server.

[1646] Step 3:

[1647] User terminal operation

[1648] Users write reviews for specific products. Text analysis technology is used to extract sentiment data from these reviews. For example, emotions such as "satisfied" or "dissatisfied" are identified based on keywords and sentence tone used in the reviews.

[1649] Step 4:

[1650] User terminal operation

[1651] The system uses a built-in camera and microphone to collect emotional data from the user's facial expressions and voice while they are browsing products or writing reviews. For example, facial recognition technology is used to determine whether the user is smiling, angry, or otherwise in a negative mood.

[1652] Step 5:

[1653] User terminal operation

[1654] The collected behavioral and emotional data is converted into a specified data format and sent to the server. The converted data includes product ID, viewing time, number of purchases, review content, and emotional state.

[1655] Step 6:

[1656] Server operation

[1657] The server receives data sent from the user's terminal. The received data is saved to the database in real time. During the saving process, data cleansing is performed, correcting inconsistent data and supplementing missing data.

[1658] Step 7:

[1659] Server operation

[1660] Using cleansed behavioral and emotional data, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, it learns characteristics that identify a user as being interested in "outdoor activities" based on past purchase history, browsing history, and emotional state.

[1661] Step 8:

[1662] Server operation

[1663] Based on a pre-trained artificial intelligence model, product recommendations are generated for the user. These recommendations take into account the user's behavioral history, emotional data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment might be suggested.

[1664] Step 9:

[1665] User terminal operation

[1666] The system receives product suggestion data sent from the server, converts it into a display format, and displays it through the user interface. Users can review the product suggestions and view details through this interface.

[1667] Step 10:

[1668] User actions

[1669] Users browse a list of suggested products, click on items they are interested in to view details, and can add items to their cart and complete the purchase if needed. After purchase, they are also required to provide reviews and feedback, documenting their feelings about the product.

[1670] Step 11:

[1671] User terminal operation

[1672] The system collects data on user purchases, reviews, and feedback. This data is converted to a specified format and sent to the server. The data includes product ID, rating score, review comments, and sentiment status.

[1673] Step 12:

[1674] Server operation

[1675] The server analyzes the response data received from the user's terminal. This analysis reveals the user's latest preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1676] Step 13:

[1677] Server operation

[1678] The AI ​​model is retrained based on the new response data. This retraining improves the accuracy of the AI ​​model, further optimizing future suggestions. New features and patterns are learned, continuously improving the model's performance.

[1679] In this way, the program collects user behavior and emotional data in real time, and uses AI to analyze and make suggestions, thereby improving the user experience.

[1680] (Example 2)

[1681] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1682] Traditional e-commerce systems only collect user behavior data to suggest products, making it difficult to provide personalized recommendations based on user emotions and interests. Furthermore, the accuracy of product recommendations was low, making it difficult to improve user satisfaction. In addition, because user emotional data was not reflected in the system, more appropriate product recommendations could not be achieved.

[1683] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[1684] In this invention, the server includes means for collecting user behavior data and emotional data, means for transmitting the collected behavior data and emotional data to the server, and means for storing and initial processing the transmitted behavior data and emotional data. This enables more personalized product recommendations by utilizing both the user's behavior data and emotional data. Specifically, it includes means for collecting emotional data in real time using user emotion recognition technology, means for inferring emotions from the user's facial expressions and voice using a camera, and means for improving the accuracy of product recommendations using the user's emotional data, thereby realizing more accurate product recommendations.

[1685] "Action data" refers to information about user actions such as browsing, purchasing, and providing feedback on e-commerce sites.

[1686] "Emotional data" refers to information about a user's emotional state obtained from facial recognition technology, text analysis, and facial expressions and voice captured using a camera.

[1687] "Means of collection" refers to hardware and software for acquiring user behavioral data and emotional data in real time.

[1688] "Means of transmission" refers to communication means for converting collected behavioral data and emotional data into the required format and sending it to the server.

[1689] "Means for saving and initial processing" refers to the process of saving the behavioral and emotional data sent to the server into a database, verifying the integrity of the data, and cleansing up any inaccurate data.

[1690] "Training methods" refer to methods for building and training artificial intelligence models that learn user preferences, behavioral patterns, and emotional patterns using stored and pre-processed data.

[1691] "Methods for generating product suggestions" refers to methods that utilize trained artificial intelligence models to generate personalized and optimal product suggestions for each user.

[1692] "Means of display" refers to a mechanism for sending generated product suggestions to the user's terminal and presenting them visually through the user interface.

[1693] "Response data" refers to information about actions users take in response to product suggestions (e.g., viewing products, purchasing them, writing reviews).

[1694] "Methods for updating" refer to methods of analyzing collected response data and retraining the artificial intelligence model to improve its accuracy.

[1695] "Emotion recognition technology" is a technology that uses facial recognition and text analysis of users, as well as a camera to analyze facial expressions and voice, in order to infer the user's emotional state.

[1696] This invention is a system that collects user behavior data and emotional data and generates product recommendations that are optimal for each individual user based on that data. This system includes the following main components:

[1697] 1. Collecting user behavior data

[1698] This refers to the user's terminal activity. This behavioral data is collected when users browse, purchase, or provide feedback on e-commerce sites. For example, if a user frequently browses specific sports equipment and purchases related products, that data is collected in real time.

[1699] 2. Collecting user sentiment data

[1700] This involves the operation of the user's terminal. It collects user emotion data using facial recognition technology and text analysis. For example, when a user writes a product review, the emotions are analyzed from the text. It also uses the camera to infer emotions from the user's facial expressions and voice. Specifically, emotion recognition software (e.g., Microsoft Azure Face API) is used.

[1701] 3. Data transmission and storage

[1702] This refers to the user's terminal activity. The collected activity and sentiment data are converted into a specified data format and sent to the server. The converted data includes details such as product ID, viewing time, number of purchases, review content, and sentiment state.

[1703] This describes the server's operation. The server stores the received data in a database (e.g., Amazon RDS) in real time and performs initial processing. This initial processing includes cleansing to ensure the accuracy of the data.

[1704] 4. Training of the artificial intelligence model

[1705] This is the server operation. Using cleansed behavioral and emotional data, it trains an artificial intelligence model that learns user preferences, behavioral patterns, and emotional patterns. For example, it runs a training script written in Python and builds the model using TensorFlow. From past purchase and browsing history, it learns that the user prefers "outdoor activities."

[1706] 5. Generating Product Proposals

[1707] This describes the server's operation. Based on a pre-trained artificial intelligence model, it generates product suggestions for the user. Here, the suggestions are made considering the user's behavioral history, sentiment data, and current trend information. For example, as summer approaches, trail running shoes and camping equipment may be suggested.

[1708] This describes the operation of the user terminal. It receives product data proposed from the server, converts it into a display format, and displays it through the user interface.

[1709] 6. Collecting user responses and feedback

[1710] This describes user behavior. Users browse suggested products, click on items of interest to view details, add them to their cart, and complete the purchase. Further data is generated when users provide reviews and feedback after the purchase.

[1711] This describes the user terminal's actions. It collects this response data, converts it to a specified format, and sends it to the server. This data includes product ID, rating score, review comments, and sentiment status.

[1712] 7. Evaluating responses and updating artificial intelligence models

[1713] This is how the server works. The server analyzes response data received from the user's terminal to understand new preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1714] This is a server operation. By retraining the AI ​​model based on new response data, the model's accuracy improves, and the next suggestions become even more optimized.

[1715] Specific example

[1716] For example, if a user who frequently browses and purchases camping gear is looking for a new tent, the system collects the user's browsing history, purchase history, reviews, and sentiment data. The server uses this data to train an artificial intelligence model and generates an optimal list of camping gear, taking into account the user's enjoyment of "outdoor activities." This list is sent to the user's terminal and displayed through the user interface. If the user purchases the suggested items and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model, further optimizing future suggestions for the user.

[1717] Example of a prompt

[1718] "Please suggest the latest tents to users who frequently browse and purchase camping equipment."

[1719] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1720] Step 1:

[1721] User behavior data collection

[1722] User terminal operation

[1723] We collect this behavioral data when users browse, purchase, and provide feedback on products on e-commerce sites. For example, when a user is browsing a specific sporting item, we record the product ID, the start time of the browsing, and the end time. If the user completes a purchase, we also collect the purchased product ID, the date and time of purchase, and the purchase amount.

[1724] Input: User's behavior on the website

[1725] Output: Operational data such as product ID, viewing time, and purchase data.

[1726] Step 2:

[1727] Collecting user sentiment data

[1728] User terminal operation

[1729] Using facial recognition technology and text analysis, we collect emotional data from users when they write product reviews or use the website. Specifically, we extract emotions from the text of product reviews written by users, and infer emotions from facial expressions and voice using a camera. For this purpose, we use emotion recognition software.

[1730] Input: User review text, camera footage, audio

[1731] Output: Emotional data such as type of emotion (positive, negative, etc.) and intensity.

[1732] Step 3:

[1733] Sending and storing data

[1734] User terminal operation

[1735] The collected behavioral and emotional data is converted into JSON format and sent to the server in real time. For example, the dataset might look like this: "Product ID: 1234, Viewing time: 5 minutes, Purchase date and time: 2023-10-12, Review emotion: Positive, Facial emotion: Smile".

[1736] Input: Behavioral data, emotional data

[1737] Output: Data in JSON format

[1738] Server operation

[1739] The server stores the received data in a database and performs a cleansing process as an initial step. This removes inconsistent or missing data to ensure data integrity.

[1740] Input: Data in JSON format

[1741] Output: Cleansed and consistent data

[1742] Step 4:

[1743] Training of artificial intelligence models

[1744] Server operation

[1745] Based on cleansed behavioral and emotional data, an AI model is trained to learn user preferences, behavioral patterns, and emotional patterns. For example, a script written in Python is executed, and the model is built using TensorFlow. The training data includes past purchase history, browsing history, and emotional states.

[1746] Input: Data after cleansing

[1747] Output: Trained AI model

[1748] Step 5:

[1749] Product Proposal Generation

[1750] Server operation

[1751] Using a pre-trained artificial intelligence model, the system generates personalized product recommendations for each user. Specifically, it considers the user's past behavioral history, emotional data, and current trend information when making recommendations. For example, "as summer approaches, it might suggest trail running shoes or camping equipment."

[1752] Input: Trained AI model, user behavior data, sentiment data, trend information

[1753] Output: Product Proposal List

[1754] User terminal operation

[1755] The system receives product data suggested by the server and converts it into a format for display in the user interface. This allows the user to visually see product images and detailed information.

[1756] Input: Product Proposal List

[1757] Output: Product list displayed in the user interface

[1758] Step 6:

[1759] Collecting user responses and feedback

[1760] User actions

[1761] Users can view suggested products, check their details, add them to their cart, and complete the purchase. After purchase, they can provide reviews and feedback.

[1762] Input: User actions (browsing, purchasing, reviewing)

[1763] Output: Response data

[1764] User terminal operation

[1765] The collected response data is converted to JSON format and sent to the server. This data includes product ID, rating score, review comments, sentiment status, etc.

[1766] Input: User action data

[1767] Output: Response data in JSON format

[1768] Step 7:

[1769] Response evaluation and AI model updates

[1770] Server operation

[1771] The server analyzes the received response data to understand new user preferences, behavioral patterns, and emotional patterns. For example, if a user gives a high rating to a particular brand of camping equipment, the server can detect a positive emotional state from that review.

[1772] Input: Response data in JSON format

[1773] Output: Updated user preference data, behavioral pattern data, emotional pattern data

[1774] Server operation

[1775] The AI ​​model is retrained based on the new response data to improve its accuracy. This will further optimize future suggestions.

[1776] Input: Updated user preference data, behavioral pattern data, emotional pattern data

[1777] Output: Improved AI model

[1778] (Application Example 2)

[1779] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1780] Traditional e-commerce systems suggested products based on users' purchase and browsing history, but it was difficult to provide optimal product suggestions that took into account user emotions and real-time behavior. Furthermore, the lack of functionality to collect and analyze user emotion data prevented personalized suggestions that reflected user feelings. As a result, improvements in the user experience were limited.

[1781] The specific processing performed 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 user behavior data and emotion data, means for transmitting the collected behavior data and emotion data to the server, and means for storing and initial processing the transmitted behavior data and emotion data. This makes it possible to provide more precise personalized product suggestions based on the user's behavior and emotions.

[1782] "Action data" refers to the user's behavioral history on the e-commerce system, including purchase history, browsing history, and feedback information.

[1783] "Emotional data" refers to information about a user's emotional state obtained through methods such as facial expressions, voice, and text analysis.

[1784] "Personalized product recommendations" refer to product recommendations optimized for each user, generated based on their individual behavioral and emotional data.

[1785] "Facial recognition" refers to a technology that captures a user's facial expressions and analyzes their emotions from those expressions.

[1786] "Text analysis" refers to the technology that analyzes text data entered by users and understands their emotions and intentions from its content.

[1787] "Trend information" refers to data about current market and consumer trends.

[1788] "Response data" refers to feedback information obtained from user actions taken in response to product suggestions (e.g., clicks, purchases, reviews, etc.).

[1789] An "artificial intelligence model" refers to a machine learning model that learns from collected and analyzed data to recognize and predict patterns and trends.

[1790] "Real-time" refers to processing and making suggestions that immediately reflect changes in user behavior and emotions.

[1791] This invention relates to a system that collects user behavior data and emotional data and generates personalized product recommendations based on that data. This system is primarily implemented using the following hardware and software components.

[1792] First, the user's smartphone will have programs installed to collect behavioral and emotional data. Collecting behavioral data includes functions to acquire information (purchase history, browsing history, feedback information, etc.) when the user browses or purchases products on e-commerce sites. Simultaneously, emotional data will be collected using facial recognition and text analysis software utilizing the smartphone's camera. Specifically, the OpenCV library will be used to capture the user's facial expressions and analyze their emotions in real time. Text analysis will use the NLTK library to analyze emotions from text data such as reviews and feedback.

[1793] The collected behavioral and emotional data are converted to a specified data format and sent to an AWS (Amazon Web Services) server via the internet. The server stores the received data in a database such as Amazon RDS and performs data cleansing using a Python script. After the data is cleansed, an artificial intelligence model is trained to learn user preferences, behavioral patterns, and emotional patterns. Machine learning algorithms such as Scikit-learn's Random Forest Classifier are used to train this model.

[1794] Next, a trained artificial intelligence model is used to generate personalized product recommendations for the user. This process takes into account the user's past browsing and purchase history, emotional state, and market trend information. The generated product recommendations are sent to the user's smartphone in real time and displayed to the user through an appropriate interface.

[1795] When a user takes action regarding a suggested product (e.g., clicking on a product, purchasing it, or posting a review), detailed response data is collected again and sent to the server. This response data is used to understand new preferences, behavioral patterns, and emotional patterns, and the AI ​​model is further updated and improved.

[1796] As a concrete example, suppose a user is browsing camping equipment and looking for a new tent. In this case, the system collects the user's browsing history, purchase history, reviews, and sentiment data. On the server side, an artificial intelligence model is trained using this data to generate a list of optimal camping equipment for the user, who is determined to prefer "outdoor activities." This list is sent to the smartphone and displayed through the user interface. If the user purchases the suggested products and posts a positive review, that sentiment data is also sent back to the server and used as training data for the AI ​​model.

[1797] Here are some concrete examples of prompt statements:

[1798] "When users write reviews about camping equipment, use review comments and facial expression data to determine their emotional state."

[1799] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1800] Step 1:

[1801] When a user browses an e-commerce site using their smartphone, the smartphone device collects operational data. This data includes the ID of the product the user viewed, the viewing time, and the number of clicks. This data is collected and stored in a local database.

[1802] Step 2:

[1803] When users write product reviews, their smartphone cameras and text input are used to collect emotional data. The input includes user facial expressions and review comments. The OpenCV library is used to analyze the facial expressions and infer the user's emotional state. Simultaneously, the NLTK library is used to analyze the review comments and recognize the emotions expressed in the text. This emotional data is also stored in a local database.

[1804] Step 3:

[1805] The smartphone device converts collected behavioral and emotional data into a specified data format and sends it to the server via the internet. The input for this process is all behavioral and emotional data stored locally. The output is the transmission process of this data. The data transmission is performed securely using the HTTPS protocol.

[1806] Step 4:

[1807] The server stores the received behavioral and sentimental data in a database such as Amazon RDS and performs initial processing using a Python script. The input consists of behavioral and sentimental data sent from the user's terminal. Initial processing includes data cleansing and formatting. The output is the cleaned data.

[1808] Step 5:

[1809] The server uses the cleansed data to train an artificial intelligence model. The inputs include cleansed behavioral data and sentiment data. A machine learning model is built and trained using tools such as Scikit-learn's Random Forest. The output is the trained AI model.

[1810] Step 6:

[1811] Based on a trained AI model, the server generates personalized product recommendations. Inputs include the trained AI model and newly submitted user behavior and sentiment data. Market trend information is also incorporated. The AI ​​model generates an optimized product list. The output is a product recommendation list.

[1812] Step 7:

[1813] The server sends the generated product suggestion list to the user's smartphone. The input is the product suggestion list generated by the AI ​​model. The output is the product suggestion displayed on the user's smartphone.

[1814] Step 8:

[1815] Action and sentiment data are collected again when the user purchases or clicks on one of the suggested products. The input is the user's new behavior and sentiment data. This data is again stored in the local database and prepared to be sent to the server.

[1816] Step 9:

[1817] The server receives new user behavior and emotion data and updates the AI ​​model. The input is the latest behavior and emotion data from the user. The AI ​​model is retrained using this new data, enabling it to make more accurate product recommendations. The output is the updated AI model.

[1818] 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 controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1819] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1820] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1821] Furthermore, the emotion identification model 59, acting 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 a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1822] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1823] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1824] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1825] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1826] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is 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 the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1827] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1828] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1829] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1830] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1831] 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.

[1832] Furthermore, it is not necessary to store the entirety 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 the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1833] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1834] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of 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). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1835] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1836] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1837] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1838] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[1839] The following is further disclosed regarding the embodiments described above.

[1840] (Claim 1)

[1841] Means for collecting user behavior data,

[1842] A means of sending the collected motion data to the server,

[1843] Means for storing and initial processing transmitted operation data,

[1844] A means for training an artificial intelligence model that learns using saved and pre-processed operational data,

[1845] A means of generating product suggestions based on a trained artificial intelligence model,

[1846] A means of displaying the generated product suggestions to the user,

[1847] A means of collecting response data from users and sending it to a server,

[1848] A system that includes means for evaluating collected response data and updating an artificial intelligence model.

[1849] (Claim 2)

[1850] The system according to claim 1, wherein the operation data includes purchase history, browsing history, and feedback information.

[1851] (Claim 3)

[1852] The system according to claim 1, further comprising means for considering trend information in generating product proposals.

[1853] "Example 1"

[1854] (Claim 1)

[1855] Means for collecting user behavior data,

[1856] A means of sending the collected motion data to the server,

[1857] Means for storing and cleansing transmitted operation data,

[1858] A means for training an artificial intelligence model that learns using saved and cleansed behavioral data,

[1859] A means of generating product suggestions based on a trained artificial intelligence model,

[1860] A means of displaying the generated product suggestions to the user through a user interface,

[1861] A means of collecting response data from users and sending it to a server,

[1862] A means of analyzing collected response data and updating the artificial intelligence model,

[1863] A system that includes means for converting data collected by a user terminal into a format and transmitting it.

[1864] (Claim 2)

[1865] The system according to claim 1, wherein the operational data includes purchase history, browsing history, feedback information, and access frequency.

[1866] (Claim 3)

[1867] The system according to claim 1, further comprising means for considering seasonal and latest trend information in generating product suggestions.

[1868] "Application Example 1"

[1869] (Claim 1)

[1870] Means for collecting user behavior data,

[1871] A means of sending the collected motion data to the server,

[1872] Means for storing and initial processing transmitted operation data,

[1873] A means for training an artificial intelligence model that learns using saved and pre-processed operational data,

[1874] A means of generating product suggestions based on a trained artificial intelligence model,

[1875] A means of displaying the generated product suggestions to the user,

[1876] A means of collecting response data from users and sending it to a server,

[1877] A means of evaluating the collected response data and updating the artificial intelligence model,

[1878] A system that includes means for displaying product suggestions to a user via a visual device used within a virtual environment, and enabling the user to select and purchase the suggested products.

[1879] (Claim 2)

[1880] The system according to claim 1, wherein the operation data includes purchase history, browsing history, and feedback information.

[1881] (Claim 3)

[1882] The system according to claim 1, further comprising means for considering trend information in generating product proposals.

[1883] "Example 2 of combining an emotion engine"

[1884] (Claim 1)

[1885] Means for collecting user behavior data and emotional data,

[1886] A means for transmitting collected behavioral data and emotional data to a server,

[1887] Means for storing and initial processing transmitted motion data and emotion data,

[1888] A means for training an artificial intelligence model that learns using saved and pre-processed behavioral and emotional data,

[1889] A means of generating product suggestions based on a trained artificial intelligence model,

[1890] A means of displaying the generated product suggestions to the user,

[1891] A means of collecting response data from users and sending it to a server,

[1892] A means of evaluating the collected response data and updating the artificial intelligence model,

[1893] A means of collecting emotional data in real time using user emotion recognition technology,

[1894] A method of inferring emotions from a user's facial expressions and voice using a camera,

[1895] A system that includes means to improve the accuracy of product recommendations using user sentiment data.

[1896] (Claim 2)

[1897] The system according to claim 1, wherein the behavioral data and emotional data include purchase history, browsing history, feedback information, and the user's emotional state.

[1898] (Claim 3)

[1899] The system according to claim 1, further comprising means for considering trend information in generating product proposals.

[1900] "Application example 2 when combining with an emotional engine"

[1901] (Claim 1)

[1902] Means for collecting user behavior data and emotional data,

[1903] A means for transmitting collected behavioral data and emotional data to a server,

[1904] Means for storing and initial processing transmitted motion data and emotion data,

[1905] A means for training an artificial intelligence model that learns using saved and pre-processed behavioral and emotional data,

[1906] A means of generating personalized product suggestions based on a trained artificial intelligence model,

[1907] A means of displaying the generated personalized product suggestions to the user in real time,

[1908] A means of collecting response data from users and sending it to a server,

[1909] A means of evaluating the collected response data and updating the artificial intelligence model,

[1910] A system including facial recognition and text analysis means for performing emotion analysis.

[1911] (Claim 2)

[1912] The system according to claim 1, wherein the behavioral data and emotional data include purchase history, browsing history, feedback information, and emotional state.

[1913] (Claim 3)

[1914] The system according to claim 1, further comprising means for considering trend information and the emotional state of users in generating product suggestions. [Explanation of Symbols]

[1915] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Means for collecting user behavior data, A means of sending the collected motion data to the server, Means for storing and initial processing transmitted operation data, A means for training an artificial intelligence model that learns using saved and pre-processed operational data, A means of generating product suggestions based on a trained artificial intelligence model, A means of displaying the generated product suggestions to the user, A means of collecting response data from users and sending it to a server, A system that includes means for evaluating collected response data and updating an artificial intelligence model.

2. The system according to claim 1, wherein the operation data includes purchase history, browsing history, and feedback information.

3. The system according to claim 1, further comprising means for considering trend information in generating product proposals.

Images